completion and initial testing of a pressurized oxy-coal
Post on 14-Apr-2022
5 Views
Preview:
TRANSCRIPT
Brigham Young University Brigham Young University
BYU ScholarsArchive BYU ScholarsArchive
Theses and Dissertations
2021-11-22
Completion and Initial Testing of a Pressurized Oxy-Coal Reactor Completion and Initial Testing of a Pressurized Oxy-Coal Reactor
Scott Hunsaker Gardner Brigham Young University
Follow this and additional works at: https://scholarsarchive.byu.edu/etd
Part of the Engineering Commons
BYU ScholarsArchive Citation BYU ScholarsArchive Citation Gardner, Scott Hunsaker, "Completion and Initial Testing of a Pressurized Oxy-Coal Reactor" (2021). Theses and Dissertations. 9290. https://scholarsarchive.byu.edu/etd/9290
This Thesis is brought to you for free and open access by BYU ScholarsArchive. It has been accepted for inclusion in Theses and Dissertations by an authorized administrator of BYU ScholarsArchive. For more information, please contact ellen_amatangelo@byu.edu.
Completion and Initial Testing of a Pressurized Oxy-Coal Reactor
Scott Hunsaker Gardner
A thesis submitted to the faculty of Brigham Young University
in partial fulfillment of the requirements for the degree of
Master of Science
Dale R. Tree, Chair Andrew R. Fry Brent W. Webb
Department of Mechanical Engineering
Brigham Young University
Copyright © 2021 Scott Hunsaker Gardner
All Rights Reserved
ABSTRACT
Completion and Initial Testing of a Pressurized Oxy-Coal Reactor
Scott Hunsaker Gardner Department of Mechanical Engineering, BYU
Master of Science Oxy-combustion is a process which removes nitrogen from air prior to combustion in order to produce a high concentration of CO2 in the exhaust. This enables CO2 liquefaction, transport, and storage to greatly reduce CO2 emissions to the atmosphere. Atmospheric oxy-coal combustion has been successfully demonstrated at industrial scales and could be retrofit in existing coal boilers, but thermodynamic efficiencies are low and therefore uneconomical. Pressurized oxy-coal combustion has the potential for higher efficiency and lower cost but requires new technologies related to the coal feed system, the burner, and ash management. This project describes work needed to complete the dry feed pressurized oxy-coal combustor (POC) at BYU. The POC required the software control system (OPTO22) to be completed, a reactor shakedown, and testing of a previously designed burner by recording reactor thermocouple, exhaust concentration, and radiometer measurements. The following has been successfully demonstrated: 1) reactor heat-up with natural gas 2) coal combustion within temperature limits of the reactor 3) slagging that allows ash management.
Keywords: pressurized oxy-coal, combustion, pulverized coal, fluidized coal feeder
ACKNOWLEDGEMENTS
I am grateful for the opportunity that I have had to study at Brigham Young University for the past six years. For the faculty, staff, and fellow students that I have interacted with during my time at this university.
I would like to thank my advisor, Dr. Tree, that mentored me for the duration of my master’s program and for all the help he has given me. I am grateful for Dr. Fry for all the hours he spent teaching and instructing. I would like to thank Dr. Adams for his encouragement and support.
I am grateful for my wonderful wife, Karisa, and the endless support she has given me throughout this process. Thank you to my parents for the encouragement and support that they have given me to pursue an education. Thank you to the rest of my family that has helped me to reach this point.
I am grateful for all the students that have contributed on this project. I would like to especially thank Dustin Badger, Dustin Barlow, and Sandra Thomas for the many hours they have contributed to this project. Thank you to the previous graduate students that laid the groundwork.
Also, I would like to thank the Department of Energy (DOE) for funding this research project.
iv
TABLE OF CONTENTS
Completion and Initial Testing of a Pressurized Oxy-Coal Reactor................................................ i
ABSTRACT .................................................................................................................................... ii
ACKNOWLEDGEMENTS ........................................................................................................... iii
TABLE OF CONTENTS ............................................................................................................... iv
LIST OF TABLES ....................................................................................................................... viii
LIST OF FIGURES ........................................................................................................................ x
1 Introduction ............................................................................................................................... 1
2 Background and Literature Review .......................................................................................... 4
2.1 Oxy-coal Combustion ...................................................................................................... 4
2.2 Atmospheric Oxy-coal Facilities and Demonstrations..................................................... 6
2.3 Pressurized Oxy-Coal Combustion Reactors ................................................................. 11
2.4 Summary ........................................................................................................................ 14
3 Component Design and Testing .............................................................................................. 15
3.1 Optical Access ................................................................................................................ 15
3.1.1 Design Requirements ........................................................................................ 15
3.1.2 Design Selection ............................................................................................... 16
3.1.3 Results ............................................................................................................... 17
3.2 Swirler ............................................................................................................................ 21
3.2.1 Design Requirements ........................................................................................ 21
3.2.2 Swirler Design and Test Results ....................................................................... 21
3.3 Purge............................................................................................................................... 24
v
3.3.1 Purge System Design ........................................................................................ 24
3.3.2 Model of Purge Gas Flow ................................................................................. 25
3.3.3 Final Design and Results .................................................................................. 27
3.4 Oxygen Sensor ............................................................................................................... 27
3.4.1 Oxygen Sensor Design Requirements .............................................................. 28
3.4.2 Design Selection and Performance Results ...................................................... 28
4 Reactor Control and Hazop Implemenatation......................................................................... 30
4.1 Hardware Requirements ................................................................................................. 30
4.1.1 Pneumatic Valve ............................................................................................... 33
4.1.2 Connecting OPTO to Building Alarm .............................................................. 33
4.1.3 Rupture Disk Tube ............................................................................................ 33
4.1.4 Ignitor ................................................................................................................ 34
4.2 Pressure Certification ..................................................................................................... 36
4.3 Software – OPTO 22 ...................................................................................................... 36
4.3.1 OPTO Control Box ........................................................................................... 37
4.3.2 OPTO PAC Control (Charts) ............................................................................ 39
4.3.3 OPTO PAC Display .......................................................................................... 40
4.3.4 OPTO GROOV BOX ....................................................................................... 44
4.3.5 POC Camera ..................................................................................................... 44
4.4 Standard Operating Procedure (SOP) ............................................................................ 44
4.4.1 Regular Maintenance List ................................................................................. 45
4.4.2 Fuel Air Ratio Calculations .............................................................................. 45
5 Results ..................................................................................................................................... 46
vi
5.1 Atmospheric Natural Gas Operation .............................................................................. 46
5.1.1 Refractory Curing ............................................................................................. 47
5.1.2 Reactor Preheating for Coal .............................................................................. 48
5.1.3 Atmospheric Natural Gas Results ..................................................................... 49
5.2 Pressurized Natural Gas ................................................................................................. 58
5.2.1 Method .............................................................................................................. 59
5.2.2 Pressurized Natural Gas Results ....................................................................... 60
5.3 Pressurized Coal ............................................................................................................. 62
5.3.1 Method .............................................................................................................. 62
5.3.2 Pressurized Coal Results ................................................................................... 65
5.3.3 Coal Flame Analysis ......................................................................................... 65
5.3.4 Coal Feeder Analysis ........................................................................................ 68
5.3.5 Ash Management Results ................................................................................. 71
6 Summary and Conclusions ..................................................................................................... 74
REFERENCES ............................................................................................................................. 77
APPENDIX A. OPTO I/O Points ........................................................................................... 79
APPENDIX B. STANDARD OPERATING PROCEDURE ................................................ 89
APPENDIX C. REGULAR MAINTENANCE LIST .................................................................. 96
APPENDIX D. MINERAL ASH, ULTIMATE, AND PROXIMATE ANALYSES .................. 98
D.1 Sample Identification: #2 ............................................................................................... 98
D.2 Sample Identification: #3 ............................................................................................. 101
D.3 Sample Identification: #4 ............................................................................................. 104
D.4 Sample Identification: #6 ............................................................................................. 107
vii
APPENDIX E. OPTO PAC CONTROL (CHARTS) ................................................................. 110
E.1 Coal Mass Flow ............................................................................................................ 110
E.2 Pressure Control Valve................................................................................................. 111
E.3 Bypass Valve Interlock ................................................................................................ 112
E.4 System Interlock ........................................................................................................... 113
E.5 Control Box Cooling .................................................................................................... 114
E.6 Emergency Trigger Logic ............................................................................................ 115
E.7 Emergency Shutdown Procedure ................................................................................. 120
E.8 Heat Loss and Inside Temperature ............................................................................... 121
E.9 Ignitor ........................................................................................................................... 123
E.10 Quench Spray VFD ...................................................................................................... 124
APPENDIX F. RECONFIGURED SUBSYSTEMS .................................................................. 125
F.1 B Thermocouples ......................................................................................................... 125
F.2 Spray Quench ............................................................................................................... 130
APPENDIX G. PRESSURIZED COAL TESTS ........................................................................ 132
G.1 First Attempt to Burn Coal ........................................................................................... 132
G.2 Second Attempt to Burn Coal ...................................................................................... 133
G.3 Third Attempt to Burn Coal ......................................................................................... 134
G.4 Fourth Attempt to Burn Coal ....................................................................................... 135
G.5 Fifth Attempt to Burn Coal .......................................................................................... 139
G.6 Sixth Attempt to Burn Coal .......................................................................................... 145
viii
Table 2-1. Summary of pilot scale oxy-coal projects in the world [15]. ........................................ 7
Table 2-2. Summary of power plant oxy-coal projects in the world [15]. ...................................... 8
Table 2-3. Atmospheric Lab Scale Oxy-Combustion Reactors .................................................... 10
Table 2-4. Pressurized Oxy-Coal Reactors. .................................................................................. 11
Table 3-1. Bill of Materials for the “half-inch” and “two-inch” window ..................................... 18
Table 4-1. HAZOP hardware list .................................................................................................. 31
Table 4-2. Ignitor bill of materials. ............................................................................................... 36
Table 4-3. The control system’s charts and the function of every chart. ...................................... 40
Table 5-1. Curing and Dryout Schedule for UltraGreen SR (Castable) Refractory. .................... 47
Table 5-2. Atmospheric natural gas flow rates with respective firing rates ................................. 50
Table 5-3. Different operating conditions tried to achieve a pressurized natural gas flame. ....... 61
Table 5-4. POC target mass flow rates for pressurized coal operation. ........................................ 63
Table A-1. List of high voltage inputs and output (HV-00-0 through HV-04-3). ........................ 79
Table A-2. List of high voltage inputs and outputs (HV-05-0 through HV-08-03). .................... 80
Table A-3. List of high voltage inputs and outputs (HV-09-00 through HV-11-03). .................. 81
Table A-4. List of low voltage inputs (LVI-00-0 through LVI-06-7). ......................................... 82
Table A-5. List of low voltage inputs (LVI-07-0 through LVI-08-7). ......................................... 83
Table A-6. List of low voltage inputs (LVI-09-0 through LVI-11-3). ......................................... 84
Table A-7. List of low voltage outputs (LVO-00-0 through LVO-09-1). .................................... 85
Table A-8. List of low voltage outputs (LVO-11-0 through LVO-15-0). .................................... 86
Table A-9. List of thermocouple inputs (TCI-00-0 through TCI-01-3). ....................................... 87
Table A-10. List of thermocouple inputs (TCI-01-4 through TCI-03-7)...................................... 88
LIST OF TABLES
ix
Table B-1. Burner MFC nominal flow rates from case 3 for a firing rate of 100 kW. ................. 95
Table B-2. Coal feed MFC nominal flow rates from case 3 for a firing rate of 100 kW. ............ 95
Table C-1. Coal Feed hardware maintenance. .............................................................................. 96
Table C-2. Reactor hardware maintenance. .................................................................................. 97
Table D-1. Mineral Ash Analysis ................................................................................................. 98
Table D-2. Ultimate Analysis ....................................................................................................... 99
Table D-3. Proximate Analysis ................................................................................................... 100
Table D-4. Mineral Ash Analysis ............................................................................................... 101
Table D-5. Ultimate Analysis ..................................................................................................... 102
Table D-6. Proximate Analysis ................................................................................................... 103
Table D-7. Mineral Ash Analysis ............................................................................................... 104
Table D-8. Ultimate Analysis ..................................................................................................... 105
Table D-9. Proximate Analysis ................................................................................................... 106
Table D-10. Mineral Ash Analysis ............................................................................................. 107
Table D-11. Ultimate Analysis ................................................................................................... 108
Table D-12. Proximate Analysis ................................................................................................. 109
Table E-1. “Emergency_Trigger_Logic” chart thresholds to decide if reactor shutdowns. ....... 117
x
Figure 3-1 Drawing of the front of the “two-inch” window housing. .......................................... 18
Figure 3-2. Drawings of the back of the “two-inch” window housing. ........................................ 19
Figure 3-3. Drawing of the “two-inch” window cap. ................................................................... 19
Figure 3-4. Drawing of the “half-inch” window holder. .............................................................. 20
Figure 3-5. Drawing of the “half-inch” window cap. ................................................................... 20
Figure 3-6. Assembly drawing of the swirler ............................................................................... 22
Figure 3-7: Testing the burner with the swirler outside of the reactor. ........................................ 23
Figure 3-8: Image through the top viewing port of the POC burning natural gas at atmospheric
pressure. ........................................................................................................................................ 23
Figure 4-1. Rendering of Reactor and auxiliary subsystems downstream with thermocouple and
pressure transducers labeled. ........................................................................................................ 32
Figure 4-2. Image of the ignitor showing the spark between the inner tube and hook on the outer
tube. ............................................................................................................................................... 35
Figure 4-3. Diagram of the OPTO Control Box. .......................................................................... 38
Figure 4-4. Middle monitor of the OPTO display. ....................................................................... 41
Figure 4-5. Left monitor of the OPTO display. ............................................................................ 43
Figure 5-1. Burner Diagram with Dimensions [8]. ....................................................................... 49
Figure 5-2. Refractory inner wall temperature calculated from the B-Thermocouples during heat-
up at a firing rate of 79.6 kW. ....................................................................................................... 51
Figure 5-3. Refractory heat loss calculated from the B-Thermocouples during heat-up at a firing
loss of 79.6 kW. ............................................................................................................................ 53
Figure 5-4. Surface magnet thermocouples placed on the reactor’s shell. ................................... 54
LIST OF FIGURES
xi
Figure 5-5. Horiba oxygen and carbon dioxide concentration measurements during heat-up. .... 55
Figure 5-6. Flue gas measurements from probe K-Thermocouple in the flue gas. ...................... 56
Figure 5-7. Surface measurements from surface K-Thermocouple. ............................................. 57
Figure 5-8. Coolant temperature measurements from probe K-Thermocouples. ......................... 58
Figure 5-9. Diagram of coal feed flows ........................................................................................ 64
Figure 5-10. Reactor refractory wall temperature at two axial locations. .................................... 66
Figure 5-11. Reactor refractory heat loss at two axial locations................................................... 67
Figure 5-12. Coal mass flow rate and three radiometer signals. ................................................... 68
Figure 5-13. One minute of radiometer signals. ........................................................................... 70
Figure 5-14. (left) zoomed out picture of the ash tank outside the reactor. (right) zoomed in
picture of the ash that has built up in the ash tank. ....................................................................... 71
Figure 5-15. An image looking up into the combustion chamber of the reactor after several hours
of burning coal. ............................................................................................................................. 72
Figure E-1. Coal feed chart flow chart........................................................................................ 111
Figure E-2. Badger valve flow chart. .......................................................................................... 112
Figure E-3. Bypass valve chart. .................................................................................................. 113
Figure E-4. Condensate ball valve chart. .................................................................................... 114
Figure E-5. OPTO control box chart........................................................................................... 115
Figure E-6. Top section of the Emergency Trigger Logic Chart. ............................................... 118
Figure E-7. Bottom section of the Emergency Trigger Logic Chart. ......................................... 119
Figure E-8. Shutdown flow chart. ............................................................................................... 120
Figure E-9. Heat and temperature calculation chart. .................................................................. 122
Figure E-10. Ignitor flow chart. .................................................................................................. 123
xii
Figure E-11. Quench spray flow chart. ....................................................................................... 124
Figure F-1. Dimensions of the POC’s combustion chamber [8]. ............................................... 126
Figure F-2. An example of the B thermocouple design used in the reactor. Two-holed ceramic
tubes spaced 50.4 mm apart. On the left end of the bottom ceramic tube an exposed bead is
visible. ......................................................................................................................................... 127
Figure F-3. Insulated wire exiting the refractory inside the “two-inch” port of the reactor. ...... 128
Figure F-4. Machined blind flange with the pressure NPT feedthroughs (on the left) and a
ceramic terminal block (inside the flange on the right). ............................................................. 128
Figure F-5. “1/8 in” NPT pressure feedthroughs (on the right) carry the B thermocouples
electrical signal outside the pressure reactor. Copper butt connectors (on the left) used to connect
the pressure feedthroughs copper wire. ...................................................................................... 129
Figure G-1. Coal mass flow, CO2 flow rates, radiometer signal and level 1 inner refractory
temperature as a function of time for pressurized coal test #4. .................................................. 136
Figure G-2. A selected period of high radiometer signal at level 1 early in the test. ................. 139
Figure G-3. Calculated coal mass flow rate during the fifth coal test with a mesh fluidization
mechanism. ................................................................................................................................. 142
Figure G-4. Radiometer voltages during the eight minute coal burn of the fifth coal test with the
mesh fluidization mechanism. .................................................................................................... 143
Figure G-5. Radiometer signal oscillations for one minute during attempt 5. ........................... 145
Figure G-6. Coal mass flow rates and radiometer signals for the duration of pressurized coal Test
#6. The data shown is averaged over 2 minutes. ........................................................................ 146
Figure G-7. Refractory inner wall temperatures calculated from B-type thermocouples during
coal feeding. ................................................................................................................................ 148
xiii
Figure G-8. Refractory heat wall calculations calculated from B-type thermocouple during coal
feeding......................................................................................................................................... 149
Figure G-9. The oscillation of the Level 1 radiometer during coal attempt 6. ........................... 150
Figure G-10. Stainless Steel connector surface K-type thermocouples during coal feeding. ..... 151
Figure G-11. Flue gas K-type thermocouples during coal feeding. ............................................ 152
1
The United States has generated almost 4 billion kilowatt-hours of electricity each year for
the last decade [1]. A large fraction of that electricity is generated with fossil fuels. The source of
fossil fuel changed dramatically over the past decade with natural gas rising from producing 24%
to 39%, while coal has dropped from 44% to 20% [1]. The EIA is currently forecasting that the
trend will reverse in 2021 and 2022 due to increases in the price of natural gas [1]. As of 2018,
electricity production by coal was the second largest source in the United States and the largest
globally [2]. Coal also benefits from huge reserves in the United States. An estimated 252 billion
short tons could last 357 years at 2019 production rates [3].
One major concern with the use of such large amounts of coal is the production of CO2 a
known greenhouse gas. Coal emits 228.6 lbm of CO2 per MBtu while natural gas emits 117.0
lbs/MBtu and Gasoline (without ethanol) produces 157.2 lbs/MBtu [4]. Promising technologies
have been proposed and demonstrated for the reduction of CO2 released from coal to the
atmosphere. One of these promising technologies is oxy-combustion. Oxy-combustion is a process
whereby fuel is burned with a mixture of CO2 and O2 instead of the normal N2/O2 mixture of air.
This is done by separating air into N2 and O2, recycling the exhaust gas from the combustion
process which contains primarily CO2 and condensing out the water at the end of the combustion
process. Without nitrogen in the system, the primary product gas of oxy-combustion is CO2 which
can be compressed to a liquid phase, easily transported in pipelines and sequestered underground.
1 INTRODUCTION
2
Oxy-coal combustion has been demonstrated to capture over 90 percent of the carbon dioxide
produced from burning coal [5].
In order for oxy-combustion to be a practical technology for the generation of electricity,
especially when compared to other fossil fuels such as natural gas, it must be efficient and most
importantly cost-effective. Oxy-combustion was initially developed as a retrofit technology on
existing coal-fired units. It has not yet been designed and optimized from the ground-up or as a
“greenfield” technology. One concept for improved efficiency and cost is to operate a pressurized
oxy-coal combustion system instead of an atmospheric system. This would provide several cost /
efficiency advantages including: 1) reducing the gas volume and decreasing the size and footprint
of a powerplant 2) reducing the parasitic load for CO2 compression at the end of the process, and
3) recovering a higher fraction of the latent heat of water in the combustion products [6].
The Department of Energy (DOE) awarded Brigham Young University (BYU) funds to
construct a 100kW pressurized oxy-coal (POC) reactor. The objective of this reactor is to design
and test technologies critical to pressurized oxy-coal combustion including: a dry coal feed system,
a high pressure burner, and high pressure ash management. The reactor design was divided into
four separate sub-systems: the fuel feed system, the burner, the combustion pressure vessel, and
the exhaust system. Each of these subsystems have been the subject of prior work including a
preliminary coal system was design by Jacob Tuia [7], the burner and combustion pressure vessel
by Cody Carpenter [8], the exhaust system by Aaron Skousen [9], a hazardous operational
(HAZOP) review by Justin Harwell [10], and a radiometer to measure axial radiative intensity by
Nicole Burchfield [11]. Detailed descriptions of these system can be found in the corresponding
masters theses.
The objective of this work is to complete and test the combined system.
3
Specific objectives include:
• Design, manufacture, install, and test incomplete subsystems including the optical
access ports, a purge system for the optical ports, a swirler for the burner, and an
oxygen sensor.
• Implement the control features identified in the HAZOP review.
• Test the burner, reactor, and exhaust system burning natural gas under atmospheric
and pressurized conditions.
• Complete commissioning tests of the entire system including the coal feeder,
burner, pressure vessel and exhaust system.
• Complete an initial coal-fired measurement of axial radiative intensity, axial wall
temperature and heat flux, exhaust gas concentration, and exhaust gas temperature.
4
This chapter will provide an overview of previous work on oxy-combustion in general and
specific information on pressurized oxy-coal combustion. Oxy-coal combustion has been a topic
of research and practical implementation for several decades while the task of performing oxy-
coal combustion above atmospheric pressure is a more recent and yet to be proven technology.
2.1 Oxy-coal Combustion
The Department of Energy (DOE) is interested in new technologies that could reduce the
amount of CO2 produced from burning coal. Proposed technologies have been grouped into three
categories: pre-combustion, post-combustion, and oxy-combustion [12].
Pre-combustion involves the removal of carbon before the combustion process is complete.
For example, a partial reaction of (coal) to carbon monoxide (CO) and then a subsequent use of
water vapor and heat to shift CO and H2O to CO2 and H2 can produce a product stream that consists
of 15-50% CO2 and the remainder being H2. The CO2 and H2 are then separated, using the H2 as a
fuel and the CO2 is captured and sequestered [13].
Post-combustion capture is the removal of CO2 following a normal combustion process. In
conventional coal powerplants, post-combustion CO2 concentrations range from 5 to 15%. The
separation of CO2 is more difficult and requires more energy at these lower concentrations and
technologies for this separation are developmental and not commercially established.
2 BACKGROUND AND LITERATURE REVIEW
5
In oxy-coal combustion, air is separated into nitrogen (N2) and oxygen (O2) prior to
combustion. The coal is then burned with a mixture of O2 and recycled flue gas (CO2 and H2O).
After condensing the water from the products, the CO2 concentration of oxy-coal combustion can
exceed 90% and can be readily compressed to a liquid using existing technologies.
Oxy-combustion has been recognized as a promising technology to reduce CO2 emissions
[14] because the high concentration of CO2 eliminates the need for separating CO2 from other
gases. It does require the process of separating O2 from N2 in the air but this process is already
commercially developed. The compression and purification unit (CPU) is required to remove
volatile components, remove other pollutants, and compress the flue gas to pipeline specifications
[5] but is less energy intensive than pre- and post-combustion CO2 capture. Concentrations of CO2
in the exhaust of over 90% have been realized [5], but this process is more costly and currently
less politically favorable than other options (natural gas combined cycle) for electrical energy
generation.
In previously demonstrated oxy-coal combustion, the pure oxygen that is fed to the burner
is obtained using a cryogenic air separation unit (ASU). This is energy intensive consuming about
0.24 kWh/kg O2 [6]. The cryogenic ASU and the CPU combine to reduce the absolute overall
efficiency of a power plant by nearly ten percent when compared to conventional coal combustion
with air [5]. Oxy-coal combustion also requires a new burner design because of the potential for a
higher adiabatic flame temperature and reduced flow velocities. Heat from the flame must be
distributed away from the burner to avoid the burner melting and allow heat to be distributed to
steam tubes [5]. Pressurized oxy-coal combustion is an attempt to reduce the inefficiencies of
atmospheric oxy-coal combustion by: 1) reducing the total volumetric flow rates of gases and
thereby enabling a reduction in equipment size and footprint, 2) decreasing the energy required to
6
pressurize and liquefy the CO2 in the exhaust gas by maintaining the inlet O2 pressure throughout
the system, 3) recovering a greater fraction of the exhaust H2O latent heat [6].
There are two possible methods for feeding coal to a pressurized reactor: dry and wet feed.
A dry feed system would transport the fuel to the reactor with non-reacting gas while a wet feed
system would mix the coal with water, producing a coal-water slurry. For higher efficiencies, a
dry feed system is potentially better because it would not require energy to vaporize the carrier
water [6]. Another advantage of dry feed systems is that they are inherently easier to control the
flame shape and length.
There are two primary challenges associated with high pressure oxy-coal burners. The first
is that the higher pressures reduce the velocity and momentum available for incoming primary and
secondary streams for a given burner firing rate. Thus, there is less energy available to mix and
stabilize flames, and less momentum to carry flame energy away from the burner where it can be
transferred to steam tubes. The second challenge is the potential for high temperatures associated
with high oxygen concentrations and the related adiabatic flame temperatures. This could lead to
high heat transfer rates near the burner, burner melting, and poor heat transfer to the desired steam
tubes in a boiler.
2.2 Atmospheric Oxy-coal Facilities and Demonstrations
Carbon capture and sequestration for coal combustion became a topic of significant interest
in approximately 2000 with the planning, building, and testing of numerous laboratory-scale, pilot-
scale and full-scale research and demonstrations projects. The Massachusetts Institute of
Technology (MIT) website for Carbon Capture and Sequestration lists full-scale power plant
projects and pilot-scale projects throughout the world as summarized in Table 2-1 and Table 2-2.
The active time for this atmospheric oxy-coal technology often referred to as first-generation oxy-
7
coal was between 2007 and 2016. During this period, two oxy-coal combustion facilities,
Schwarze Pumpe and Callide-A, successfully demonstrated the combined technologies of oxygen
separation, O2/CO2 oxidization of coal, and CO2 capture and conversion to a liquid phase. A third
project, Compostilla, successfully demonstrated oxy-combustion of coal in a fluid bed and carbon
capture. Conversely, although there have been numerous projects aimed at full-scale power plant
demonstrations of oxy-coal combustion, they have all been cancelled prior to completion. The two
factors reported on the MIT website for numerous cancellations are costs and political climate.
While the cost of oxy-coal combustion has not increased from projected values, the cost of
electricity production has decreased due to lower than expected natural gas prices. Alternative
methods for CO2 reduction through renewable energy sources has also decreased in cost making
oxy-coal combustion less attractive.
Table 2-1. Summary of pilot scale oxy-coal projects in the world [15].
Pilot Scale Project Name
Location Description Time Period Outcome
Schwarze Pumpe
Vattenfall, Germany
30 MW 2007 – 2014 Operational for 18 months 2009-2010. Complete operation, production of steam and nearly 100% CO2 capture. Cancelled 2014 cost
Compostilla Ponferrada, Spain
30 MW 2009 – 2012 Completed 30 MW fluid bed test. Cancelled larger scale plans 2013.
Callide-A 30 MW 2012 - 2015 2 yrs 9 months, 70 T/day of demonstrated oxy-coal with 90% capture rate.
8
Table 2-2. Summary of power plant oxy-coal projects in the world [15].
Power Plant Project Name
Location Description Time Period Outcome
Future Gen Illinois, USA 200 MW , Coal Power Plant
2014-17 Reaching planning and permitting phase, Cancelled 2015
Kimberlina California, USA
50 MW, Coal and Biomass Power Plant
2011 – 2015 Carbon capture demonstrated 2004. Project cancelled before power plant built
Compostilla Ponferrada, Spain
320 MW 2009 – 2012 Cancelled larger scale plans 2013.
Janschwalde Brandenburg, Germany
300 MW 2008 – 2016 Feasibility study and plan completed, cancelled 2011
White Rose North Yorkshire, UK
426 MW 2014 – 2020 Cancelled in 2015 before construction
Korea CCS 500 MW 2010 – 2020 Unknown – Web site not updated after 2016
The full- and pilot-scale oxy-coal combustion projects have been supported by numerous
laboratory scale combustion projects that have been used to study some of the unique challenges
of oxy-coal compared to air-fired coal combustion. Seven lab scale atmospheric reactors and their
respective sizes are listed in Table 2-3. These reactors vary in size from 20 to 1000 kWth. Every
reactor mentioned has been built and demonstrated to feed and burn coal in oxygen rich conditions.
The Brigham Young University down-fired reactor operates at a firing rate of 150 kW.
During operation, ash is collected in a cyclone located at the reactor exit. The cyclone collects ash
particles of 2 microns and larger. Three burner configurations were investigated with variable
amounts of oxygen flow in the center tube. Average intensity, temperature profiles, NO
concentrations, and loss of ignition (LOI) have been reported. This reactor explored the effects of
injecting oxygen into the near burner region of a swirl-stabilized biomass burner. It was found that
at low oxygen flow rate of 2 kg/hr two-flame regions were present and upon increasing the flow
9
rate to 8 kg/hr the two-flame region was less distinct, and the flame became elongated and more
distributed. This work characterized the flame produced from two hardwood biomass fuels burning
under oxygen rich conditions.
The TU Darmstadt 20 kW reactor has a swirl-stabilizing burner. Three operating conditions
have been investigated. Air operation was used as a reference case and oxygen concentrations of
25% and 30% have been explored. The inert gas introduced with the oxygen streams is CO2.
Images of the flame inside the quarl including visible and OH PLIF were captured for the three
different operating conditions. A quarl is frequently used to stabilize turbulent flames and this
work explored the effect of oxygen concentrations on the flame shape inside the quarl region.
The RWTH Aachen University down-fired swirl flame reactor is 60 kW. This work
explored the effect of different oxygen concentrations in the streams being introduced. The three
operating conditions published by this reactor were 1) an air firing base case 2) a 21% (vol.) oxygen
to CO2 ratio and 3) a 25% (vol.) oxygen to CO2 ratio. Flame radiation was measured using a
narrow-angle spectrometer (300-1000nm) and a cold-background probe. Images were also taken
of the combustion chamber through an observation port. The radiation measurements helped
characterize the temperature profile with the reactor during air and oxy-fuel combustion.
The University of Utah’s oxy-fuel combustor (OFC) is a 40 kW thermal load capacity
down fired reactor. The burner consists of a central primary tube and an annular secondary flow
with fixed 30o swirl vanes. Quartz glass windows allow for optical access to the flame region.
Purge gas was introduced to keep the quartz glass windows clean during operation. Changes in
oxidizer flow rate visibly changed flame luminosity, flame liftoff, and flame length but the flame
remained elongated and detached for all operating conditions.
10
The University of Utah’s 1500 kW pilot scale furnace can operate under air and oxy
conditions. The burner has a coal pipe, inner secondary, and outer secondary registers like a
traditional low-NOx burner. The burner’s oxygen may be premixed with recirculated flue gas
(RFG) at independent O2 concentrations in each register. Each register was sized to provide the
desired gas velocity range with either air, or a mixture of O2 and RFG.
The University of Leeds’ lab scale down-fired reactor is 20 kW. A swirl-stabilizing burner
supplies a swirled secondary stream and a non-swirled primary air stream. A port also allows for
over-fire air with oxygen injection. Cases have been performed with varying oxygen
concentrations in the secondary stream, varying amounts of staged air, and varying oxygen
concentrations in the staged air. During these cases, NO, CO2, and O2 concentrations have been
measured. Additionally, temperature measurements at the burnout zone and near the burner have
been published. This study concluded that oxygen rich environments result in a 1.1% to 1.5%
better carbon burnout and a measurable decrease in the production of NO during oxygen
enrichment processes.
Table 2-3. Atmospheric Lab Scale Oxy-Combustion Reactors
Reactor Location Reactor Size (kW)th
Reference
Brigham Young University (BYU) 150 [16] TU Darmstadt (Germany) 20 [17] RWTH Aachen University (Germany) 60 [18] Institute of Engineering Thermophysics, Chinese Academy of Sciences (IET/CAS) (China)
1000 [19]
University of Utah 40 [20] University of Utah 1500 [21] The University of Leeds (United Kingdom) 20 [22]
11
2.3 Pressurized Oxy-Coal Combustion Reactors
To reduce the cost of oxy-coal combustion, pressurized oxy-coal is being investigated. At
the time of this writing, only two laboratory scale pressurized oxy-coal reactors and no pilot- or
full-scale pressurized oxy-coal reactors have been reported as operational in the literature. The two
oxy-coal reactors are listed in Table 2-4. Both coal reactors discussed in this section have operated
at pressures of fourteen to fifteen bars. They have demonstrated feeding and burning coal under
oxygen rich conditions.
Table 2-4. Pressurized Oxy-Coal Reactors.
Reactor Location Reactor Size (kW)th Coal Feed System Reference University of Utah (USA) 300 Slurry [23] Washington University in St. Louis (USA)
100 Dry [24]
A staged 100 kWth pressurized oxy-combustion facility was designed and constructed at
Washington University [24]. The down fired reactor is 3 meters in length and 127 mm in diameter.
The reactor consists of a furnace contained in the center of a pressure vessel. CO2 purge gas
between the furnace and pressure vessel cools the pressure vessel shell. During operation oxygen
concentrations at the outlet, wall temperatures at multiple locations, and particle size distributions
have been measured. The reactor has operated at four conditions with a thermal input of 50,75,
100, and 125 kW of coal.
The burner designed for the pressurized application consists of a coaxial design. The burner
consists of a center tube, inner annulus, and outer annulus. Oxidizer is introduced through the
center tube and outer annulus and coal is introduced through the inner annulus. A flame holder is
located at the exit of the burner on the outside of the fuel tube. The burner was designed to produce
a low-mixing flow field, resulting in a long and straight flame to distribute the heat release along
12
the length of the furnace. To accomplish this the reactants are not swirled. Coal is introduced to
the burner by a pressure vessel hopper filled with coal. The hopper is filled prior to each operation
at atmospheric pressure. After the hopper is full it is closed, and the furnace and hopper are
pressurized.
The combustion chamber of the reactor is comprised of five sections. At the end of the
burner a quartz tube allows for visualization of the flame. A water-cooled wall with a small
opening for a high-speed camera cools the quartz section of the reactor. After the quartz section,
four refractory insulated sections with six access ports at three different heights make up the rest
of the combustion chamber. Following the combustion chamber quench water and a water bath
cools the flue gas. The reactor’s ash is managed by the quench water and a PM10 cyclone. Most
ash particles in the flue gas are washed out by the spray and fall into a water bath is reported by
Yang [24].
Washington University has concluded that the pressurized oxy-combustion technology can
produce a minimum of 3.5% higher thermodynamic efficiency than conventional oxy-combustion
plants [24]. Tests results show that stable coal flames could be achieved with no gaseous fuel
support. It was also found that complete coal combustion can be achieved with as low as 0.8%
oxygen mole fraction in the flue gas. Future studies will characterize ignition, heat transfer, SOx
and NOx formation, ash formation and partitioning.
The second lab scale pressurized oxy-combustion facility listed in Table 2-4 has been built
and demonstrated at the University of Utah. The pilot-scale, refractory-lined, down-fired reactor
supports firing rates up to 300 kW at a maximum of 14 bar. The reactor is 1520 mm long and 203
mm in diameter. Coal was fed as a coal-water slurry and operation was very sensitive to the ratio
13
of water and coal fed because the ratio of water and coal fed was the only way to adjust the
temperature of the system.
The burner consists of a center tube, inner annulus, and outer annulus. The center tube
introduces secondary O2, an inner annulus introduces the fuel, and the outer annulus delivers the
primary O2. The primary O2 is forced through a relatively small annulus creating high O2
velocities. The relatively high velocities with respect to the coal-water slurry (CWS) at an angle
of 30o from the CWS. These conditions atomize the CWS. The secondary O2 is used to improve
the atomization process similarly to effervescent atomizers.
The data collection consisted of B thermocouples, radiometers, and a gas analyzer. Four
pairs of B thermocouples placed at different levels of the combustion region provided wall heat
flux and inner wall temperature measurements. These B-thermocouples embedded in refractory
were spaced a known distance apart and a known distance from the inner wall of the refractory.
One side of the reactor features CaF2 ports to allow for the optical measurements of the
radiometers, while the other side of the reactor. Nitrogen purge gas can be introduced at each
window presumably for window cleaning purposes. These radiometers gathered heat flux
measurements during operation.
The University of Utah has operated their pulverized slurry fed Utah bituminous coal at
temperatures up to 1600 oC and pressures from 6.7-14 bars in an oxygen-rich environment. It was
concluded that operation was very sensitive to the ratio of water and coal fed. The reactor
temperature significantly impacted the carbon burnout efficiency and that the sulfur mostly
remined in the ash. The successful experimentation demonstrated that pressurized coal combustion
is a viable combustion method.
14
In addition to these two reactors. a 5 MW flameless pressurized oxy (FPO) pilot scale
reactor has been operated for more than 18,000 hours [25]. However, no testing results have been
published and only a conference presentation could be found to provide information about this
reactor. This reactor operates at 4.63 bars and uses a slurry coal feeding mechanism. A dry feed
pressurized oxy coal reactor has also been designed and constructed at the University of Utah, but
no testing results have been published.
2.4 Summary
Pressurized oxy-coal combustion is a new technology where little experimental data has
been collected. Only two laboratory scale reactors that have been built and operated to date. A
pressurized oxy-coal system at BYU will help further investigate this new promising technology
for reducing CO2 emissions. Furthermore, the dry coal feed system at BYU is unique because the
Washington University reactor does not transport the coal via fluidization and the University of
Utah reactor transported its coal as a slurry. Valuable information for characterizing pressurized
oxy-coal combustion is still needed to understand the viability of pressurized oxy-coal combustion.
15
Although the majority of the POC had been designed and fabricated prior to this work,
several components still needed to be designed, fabricated, and tested before measurements could
be obtained. These components included high pressure windows, a burner swirler, a purge system,
and an oxygen sensor.
3.1 Optical Access
Optical access to the POC is required to obtain radiation and transmission measurements.
Two different types of radiation measurements; one collecting total radiation with a radiometer,
and a second collecting spectral radiation using an FTIR are needed. Also, transmission
measurements using multiple wavelengths from an argon ion laser are planned.
3.1.1 Design Requirements
The viewing ports need to hold windows that withstand twenty atmospheres of pressure
and temperatures up to 230 oC. The radiation measurements require high optical transmission over
a wide range of wavelengths. The reactor was designed with twenty access ports; four ports
positioned 90 degrees apart at five different levels or axial positions. Ten of these ports, two on
each level, 180 degrees apart are designed to allow optical access with a line-of sight through the
3 COMPONENT DESIGN AND TESTING
16
reactor. These ports are 5.08 cm (2 in) cylindrical bores through the insulation with “two-inch”
flanges on the end. The optical access ports were required to fit onto these existing access flanges.
In addition to these “two-inch” viewing ports, “half-inch” flanges are located on two
viewing ports in the reactor cap for mounting UV sensors that monitor the presence of a flame. A
different window was designed for mounting on these smaller flanges. In addition to the geometric
constraints, these smaller windows must accommodate a “half-inch” NPT nipple to be welded onto
the window housing; this nipple enables the UV sensors to be threaded directly onto the housing.
3.1.2 Design Selection
Sapphire was selected for the window material because of its superior strength and
transmission capabilities. A 10 mm thick sapphire window transmits with greater than 80%
efficiency from 300 to 5000 nm. Quartz transmits at greater than 80% over a similar spectral range
as sapphire (from 200 nm to 2500 nm) but has a significant drop in transmittance between 2500
and 3000 nm which could negatively impact spectral radiation measurements and total radiation
at lower temperatures. Sapphire also has a significantly higher rupture modulus allowing for a
thinner window. As a result, sapphire was selected as the window material.
The thickness of the window must be determined for the windows to operate safely at
twenty atmospheres. The modulus of rupture can be used to find the thickness as a function of
pressure. Equation (3-1) shows the thickness as a function of pressure (P), area (A), rupture
modulus (𝜎𝜎𝑟𝑟), and safety factor (S). The sapphire window properties used in the calculations were
found online from Mt-Berlin [26]. With a rupture modulus of 280 MPa and a safety factor of 2 the
minimum thickness calculated for 1924 kPa (280 psi) at 500 oC was 5.1 mm (0.20 inches). A 6.35
mm (0.25 inch) sapphire window was commercially available and was purchased for testing.
17
𝑡𝑡 = 𝑆𝑆�𝑃𝑃𝑃𝑃
3.12𝜎𝜎𝑟𝑟 (3-1)
3.1.3 Results
SolidWorks was used to model the geometry of the window housing and the results are shown in
Figure 3-1 through Figure 3-5. The housing was designed to modify commercially
available 300# blind flanges. A flat bottom hole was machined into the backside of the blind flange
for the window to rest. This blind flange will be called “window holder”. Additionally, an O-ring
groove was machined into the flat bottom hole. A window cap was machined out of stainless steel
stock material. The purpose of the window cap is to hold the window in place and compress the
O-ring in its groove. The window cap mounts to the window housing by a screw pattern machined
into the window holder that allows screws to hold the two pieces together. A separate O-ring is
compressed between the window and the window cap to prevent damage to the window. A list of
18
the O-rings for both the “two-inch” and “half-inch” flanges are contained in Table 3-1. All twelve
windows were hydrostatically tested up to 2.8 MPa, gauge pressure, prior to being mounted on the
reactor; none ruptured during testing.
Table 3-1. Bill of Materials for the “half-inch” and “two-inch” window
Part Material Supplier Item # “Two-inch” Groove O-ring (#031) Viton McMaster-Carr 9464K311 “Two-inch” Cap O-ring (#031) High-Temp Silicone McMaster-Carr 1182N031 “Two- inch” Flange Screw Stainless Steel McMaster-Carr 92196A301 “Two-inch” Window Lens Sapphire Swiss Jewel W51.00 “Half-inch” Groove O-ring (#015) High-Temp Silicone McMaster-Carr 1283N26 “Half-inch” Cap O-ring (#015) High-Temp Silicone McMaster-Carr 1182N015 “Half-inch” Flange Screw Stainless Steel McMaster-Carr 92196A198 “Half-inch” Window Lens Sapphire Swiss Jewel W19.25
Figure 3-1 Drawing of the front of the “two-inch” window housing.
19
Figure 3-2. Drawings of the back of the “two-inch” window housing.
Figure 3-3. Drawing of the “two-inch” window cap.
20
Figure 3-4. Drawing of the “half-inch” window holder.
Figure 3-5. Drawing of the “half-inch” window cap.
21
3.2 Swirler
During initial testing of the burner at atmospheric conditions it was determined that the
natural gas flame was unstable. A swirler which increases the angular momentum of the fuel or
oxidizer and creates a recirculation zone was needed to stabilize the natural gas flame. As a
result, a swirler was designed for the air flow through the outer annulus of the burner previously
designed by Carpenter [8]. This also increases the mixing and results in a lean flame that is blue
in color and emits UV radiation visible to the flame detectors. The swirler does not affect the
coal flame which does not utilize the outer annulus.
3.2.1 Design Requirements
The natural gas flow to the reactor burner is controlled and hardware locked by UV sensors.
The ultra violet (UV) sensors detect UV emission or a blue flame created from excited CH radicals.
The swirler mixes air into the fuel which reduces the formation of soot which can mask CH radical
emission. The swirler must fit and be installed into the outer annulus which has an outer diameter
of 28.5 mm (1.12 in) and the inner diameter is 9.53 mm (0.375 in). The swirl vanes should produce
maximum possible swirl for a given flow rate in the burner without falling out of the burner during
operation.
3.2.2 Swirler Design and Test Results
A drawing of the final design is shown in Figure 3-6. The swirler is comprised of five
stainless steel parts. The body is machined from stainless steel, and initially an external and internal
groove were designed for O-rings. Unfortunately, the part would not fit in the burner with the O-
rings installed and the O-rings were never used. Thermon, a commercially available high
22
temperature adhesive was applied instead to keep the swirler in place and seal along the outer
circumference.
To create swirl, four holes 90 degrees apart were drilled through the swirler. Four 90 degree
Union Elbow Micro-Fit (Miniature Tube Butt Weld) Fittings, Swagelok product number 6LV-
4MW-9, were welded onto these holes to swirl the air. Each Union Elbow comes with two butt
weld connections. For the four Union Elbows to fit in the burner, one of the two butt weld
connections of each Union Elbow was removed.
Figure 3-6. Assembly drawing of the swirler
The burner, with installed swirler, was tested outside of the reactor. In Figure 3-7, a blue
flame can be seen near the exit of the burner followed by a diffusion flame. Although the luminous
soot appears to dominate emission, the UV sensor was able to detect the flame during testing.
23
Figure 3-7: Testing the burner with the swirler outside of the reactor.
Further testing was done with the burner installed in the reactor. A blue flame was visible
through the viewing ports as seen in Figure 3-8. The UV sensors were able to detect the flame
and the flame was stable for an indefinite period. After numerous startup procedures, the natural
gas flame still appears as it did in the initial startup suggesting the swirler is still in place.
Figure 3-8: Image through the top viewing port of the POC burning natural gas at atmospheric
pressure.
24
3.3 Purge
The high pressure sapphire windows were designed for taking radiative emission
measurements. To avoid condensation and particle deposition on the windows and in order to avoid
absorption of radiation in the passages between the hot product gas and the window, a purge gas
was needed. The 50.8 mm (2 in) diameter passages between the combustion chamber and the
sapphire windows are 43.18 cm (17 in) long and produce a path where cold gases and particles
may attenuate radiation from combustion products. The purge gas needs a high enough flow to
prevent fine particles from moving down the passage and depositing on the walls, must have a
high transmittance for radiation, and have a minimal impact on combustion reactions. Additional
considerations for the purge gas are the costs and availability and the amount of time gases can
flow continuously. In total, there are two UV sensors, ten viewing ports, and a rupture disk that
need to be purged. Tubing must be installed from the supply to all thirteen locations. The system
must purge during the entirety of the POC’s operation.
3.3.1 Purge System Design
To ensure that all thirteen locations have the same mass flow rates, a system was designed
with critical flow orifices sized to choke the flow. Nitrogen was selected for initial testing because
nitrogen does not negatively impact the results. In cases where the measurement of nitrogen oxides
are critical for drawing conclusions, argon can be used in place of nitrogen. A single line feeding
all windows are connected to two gas cylinders. A valve allows the gas source to be switched from
one cylinder to another allowing continuous operation and replacement of cylinders.
25
3.3.2 Model of Purge Gas Flow
Suda and Yamada [27] found that the velocity of the purge gas needed to keep water vapor
from diffusing and condensing on cold surfaces was 2 mm/s. Using a safety factor of 2.5, a 5 mm/s
velocity in the ten “two-inch” ports was targeted. The purge gas flow was modeled as a
compressible fluid as presented in chapter three of Gas Dynamics by James and Theo [28]. The
gas flow is assumed to be one-dimensional and steady-state flow. Mass flow rate in the reactor’s
“two-inch” port can then be modelled by Equation (3-2).
��𝑚 = ρAu = constant (3-2)
Compressible flow is considered when modeling the mass flow rate through the orifices,
since the orifices were designed to operate under choked conditions. Choke flow is achieved
when the pressure conditions as shown in Equation (3-3) are met. Where P is the static pressure,
and Po is the stagnation pressure.
𝑃𝑃𝑃𝑃𝑜𝑜≤ 0.52828 (3-3)
By assuming an ideal gas with constant specific heats Equation (3-4) can be used to
determine the mass flow rate through the orifice.
��𝑚 =
𝑃𝑃𝑅𝑅𝑅𝑅
𝑀𝑀�γRTπd2
4
(3-4)
Where P is pressure, T temperature, R the ideal gas constant, M is the Mach number, γ
the ratio of specific heats, and d is the orifice diameter. Setting Equations (3-2) and (3-4) equal
26
and solving for the orifice diameter results in Equation (3-5). It is assumed that M is unity due to
the choked flow condition.
d = �4�RTγρAuπP
(3-5)
With this derivation, a diameter of 0.12 mm was calculated to provide 5 mm/s of purge gas
flow in the 50.8 mm diameter viewing port region, while supplying gas to thirteen different
locations with a pressure, P, of 483 kPa (70 psi). With the ideal gas law, it was also calculated how
long it would take for a nitrogen k-cylinder to empty. The compressed nitrogen gas cylinder’s
internal volume is 49.9 L. With the mass flow rate calculated above and the initial pressure of 13.7
MPa (2000 psig), the time needed to empty the cylinder is calculated with Equation (3-6).
t =
(ΔP)VmRT
(3-6)
The purge was calculated to last 20 hrs when supplied by a 49.9L, 13.7 MPa cylinder and
the upstream and downstream pressures of 481 kPa (70 psi) and 100 kPa (14.5 psi), respectively.
At twenty atmospheres, the gas is denser by a factor of twenty and requires approximately
twenty times the flow rate to achieve the same velocity. Also, the minimum pressure that the
cylinder must provide to keep the flow choked increases to approximately 4.1 MPa (600 psig).
Therefore, the time taken to empty a tank is reduced by more than a factor of twenty.
27
3.3.3 Final Design and Results
The purge system has been installed and can deliver gas to all thirteen locations from two
k-cylinders at one time. An orifice from O’Keefe Controls Co., Part number D-6-SS, with a
diameter of 0.150 mm (0.0059 in) was installed at all thirteen locations. The gas reservoir consists
of two k-cylinders with a combined 11,328 L (400 scf) at 13.7 MPa (2000 psig). By providing
purge gas to the windows while a flame was in the reactor, the pressure upstream of the orifices
was decreased until water started to condense on the windows. With the reactor at atmospheric
pressure, it was found that at 689 KPa (100 psig) condensate started to form on all windows. The
gas velocity in the 50.8 mm ports for these conditions was calculated to be 14 mm/s. This velocity
is seven times more than what was found by Suda and Yamada [27]. The time found to empty two
tanks with the reactor at atmospheric pressure was twelve hours. With four tanks, a maximum run
time of twenty-four hours may be achieved. The purge flow has worked effectively during heat up
with natural gas but when coal is introduced, a surge of coal causes deposition on the windows. At
1.5 MPa, the purge has been operated with a pressure of 3.4 MPa (500 psi) upstream of the orifice.
With these conditions, the purge has been noticed to decrease at a rate of 1.7 MPa (250 psi) per
hour. The current system can do nothing to clean the windows after deposits have formed and
cannot eliminate deposition when the coal feed pulses dramatically.
3.4 Oxygen Sensor
An oxygen sensor is imperative to the safe operation of any combustion reactor. The
oxygen sensor allows the operator to know if the fuel entering the reactor is oxidizing. If fuel is
continually introduced but not oxidized a dangerous and undesirable energetic event may occur.
The oxygen sensor also provides information to determine coal and oxygen flow rates which may
be more accurate and responsive than the mass flow measurements.
28
3.4.1 Oxygen Sensor Design Requirements
The more robust and the faster the time response of the oxygen sensor, the more useful it
will be in operating the reactor. In order to have a fast time response, the oxygen sensor must be
robust enough to operate in the hot flue gases reliably. The sensor must be installed as close to the
combustion chamber as possible to reduce the residence time. The sensor must operate with the
other product gases present including water vapor and SO2 and it must be tolerant to ash particles.
3.4.2 Design Selection and Performance Results
A zirconium oxygen sensor, as is typically used to measure oxygen in the exhaust of a
spark ignition engine, was selected because of its rapid response time (less than 1 second) and the
ability to operate in hot product gases. The Bosch Lambda Sensor LSU 4.9 was selected to be used
for this application because of positive previous experiences with this sensor. This sensor measures
lambda (λ) which is the inverse of the equivalence ratio. The sensor was purchased from Innovative
Motorsports, part number 3888, and an LC-2: Digital Wideband “Lamda” O2 Controller. This
sensor is able to measure a λ of 0.65 to 8.0. A LC-2 user manual is located currently on the group
folder at /doe-poc/Reference Material/O2 Sensor/LC-2_Manual_O2_Sensor.pdf. This manual
contains information on calibration procedures, error codes, and troubleshooting tips.
The LC-2 has two separate output signals and both are 0 to 5 volts. One of these signals is
input into an OPTO module that supports this signal type. In the user manual it specifies that the
0 to 5 volt signal is linear; however, from testing it appears that the signal is parabolic. A zirconium
sensor must be installed at atmospheric pressure to provide an accurate reading. The controlling
factor in residence time is distance from the sensor to the reactor. To shorten the response time,
the sensor was installed directly downstream of the pressure control valves. During operation a
residence time of a 2 – 3 seconds was observed.
29
The O2 sensor performed reasonably well but it was found that liquid water impingement
on the sensor while it was operational would damage it permanently. When the gas flow rate
through the exhaust systems exceeds 124.3 kg/hr, it has been observed that the quench water
becomes entrained in the flow and transports liquid water through the heat exchanger and past the
pressure control valves, impinging water on the sensors. The work-around for this problem has
been to operate the reactor during heat-up below 80 kW where total flow rates are low enough to
avoid water entrainment.
30
A Hazard and Operability Study (HAZOP) was performed to identify potential safety risks
when operating the POC. The HAZOP committee consisted of BYU safety officers, Chemical
Engineering and Mechanical Engineering faculty, and Justin Harwell, a graduate student who was
assigned the development of a HAZOP as part of his MS thesis. The findings of this study were
documented by Harwell [10]. The HAZOP identifies three categories: hardware requirements,
software requirements, and standard operating procedure. The following chapter describes the
implementation of the HAZOP study.
4.1 Hardware Requirements
This section discusses the hardware that was added to the POC and auxiliary system
following the HAZOP study. Table 4-1 contains a list of HAZOP ID numbers, descriptions, and
reasons identified by the HAZOP for each piece of hardware that was installed to satisfy the
HAZOP requirements. The HAZOP ID refers to identification numbers in Harwell’s report [10].
The hardware listed was either added or repurposed as a result of the HAZOP. The list contains
six pressure transducers or sensors, five check valves, two sets of thermocouples, two flow meters
and several other items. Every piece of hardware on this list has been installed, wired, software
connected, and tested.
4 REACTOR CONTROL AND HAZOP IMPLEMENATATION
31
Table 4-1. HAZOP hardware list HAZOP ID Hardware Comments/Explanation
52 Quench Spay Solenoid
Replaced the quench spray solenoid with a fail open such that in the case of a power outage the spray will continue to flow to prevent overheating.
53 Emergency Drain Valve.
Adjusted the emergency drain valve to be fail close. In the case of a power outage the valve would close containing the gases in the reactor instead of releasing them uncontrollably.
56 Pressure Relief Valve
Installed a pressure relief valve for the heat exchanger coolant line. In the case of the coolant line failing and being open to the reactors pressure; a pressure relief valve will prevent the line from pressurizing.
69 Oxygen Supply Line Check Valve
Installed a check valve on the oxygen supply line. Stops backflow into the oxygen supply line.
1 Room Alarm Lights
Installed alarm lights above the ERL 106C door. The stack light has red, yellow, and green; the lights will illuminate to communicate to individuals in the ERL 106 the state of the reactor.
62 High Pressure Natural Gas Check Valve
Installed a check valve on the high pressure natural gas supply line. Stops backflow into the high pressure natural gas supply line.
63 Building Alarm System
Tied in the building alarm system into OPTO. OPTO receives a signal when the building alarm is on.
24 Rupture disk’s alumina tube
Installed an alumina tube protruding from the port that connects the reactor to the rupture disk. Stops slag from building up over the port effectively sealing off the rupture disk from the reactor.
66 High Pressure Air Check valve
Installed a check valve on the high pressure air supply line. Stops backflow into the high pressure air supply line.
23 Reactor Surface Thermocouples
Wired and installed reactor surface thermocouples. Magnet thermocouples monitor the reactor’s shell temperature.
67 High and Low Pressure Transducers
Wired and installed pressure transducers on the high and low pressure air lines. The high and low pressure air lines control pneumatic valves that require 80 psig. to function properly.
2 & 68 Oxygen Sensor Installed an oxygen sensor. Provides oxygen concentration measurements of the products of combustion.
55 Pressure transducers
Installed three pressure transducers to calculate pressure differentials within the reactor. Pressure differentials are used to tell if corrosion build up in the reactor are restricting flow or are causing blockages.
32
76 Coolant line flow meter
Installed and wired a flow meter on the heat exchanger coolant line. Needed to calculate the heat removal in the heat exchanger.
51 Quench spray flow meter
Wired a flow meter on the quench spray water line. Allows the water being put into the system to be measured.
70 Pressure differential sensors
Installed and wired pressure differential sensor for the cyclone and dilution damper. A negative pressure is needed to stop products of combustion from entering the room.
16 Rupture disk thermocouple
Wired and installed a thermocouple downstream of the rupture disk. Monitors if the rupture disk has burst.
An image of the reactor with the location of thermocouples and pressure transducers is
shown in Figure 4-1. Most of these pressure transducers and thermocouples were installed due to
the HAZOP study.
Figure 4-1. Rendering of Reactor and auxiliary subsystems downstream with thermocouple and
pressure transducers labeled.
33
Details for some of the components listed in Table 4-1 are provided in the following
subsections.
4.1.1 Pneumatic Valve
The pneumatic valve, HAZOP ID# 53, was changed from fail open to fail close. This
pneumatic valve is used to slowly vent the gases within the reactor in the case of an emergency.
The HAZOP review concluded that in the case of a power outage, the hot products of combustion
should remain in the reactor. To do so the position of this valve needed to be changed. To change
its failure position, the pistons within the valve were removed and the initial position of the pistons
were adjusted manually. Detailed instructions on the process are on the manufacturer’s website.
With this valve adjusted to be fail closed, the hot products gases will not leak into the exhaust duct
in the event of a power outage.
4.1.2 Connecting OPTO to Building Alarm
The building is equipped with an emergency alarm system that detects harmful gases such
as carbon monoxide and sulfur dioxide. These alarms were connected to OPTO using the relay
located in the junction box above the ERL 106C door. This relay sends out a high voltage signal
to the OPTO high voltage controller when the building alarms are on. With the alarm system as an
input, the alarm state can be used with OPTO to control reactor functions.
4.1.3 Rupture Disk Tube
An alumina tube was installed through the pipe that connects the rupture disk to the reactor.
This tube protrudes from the port to allow slag to flow around the tube. Ideally, slag will flow
around the tube to reduce the chance of slag hardening over the port. This precaution was taken to
34
reduce the chance of the rupture disk being sealed off from the reactor. The alumina tube is a
slightly smaller outer diameter than the inner diameter of the port to allow for easy installation.
Because of the high temperatures, there has been no material found to seal the tube into place.
4.1.4 Ignitor
The ignitor consists of three tubes that fit concentrically within each other. An image of
the current ignitor is shown in Figure 4-2. The components are outer and inner stainless steel
electrically conductive tubes and a ceramic electrically insulative alumina tube in the center. The
outer stainless-steel tube has an outer diameter of 12.7 mm (0.50 in) and a wall thickness of 1.3
mm (0.05 in) or inner diameter of 10.2 mm (0.40) in. The alumina has an outer diameter of 9.5
mm (0.38 in) and an inner diameter of 6.4 mm (0.25 in). The inner tube has a 6.1 mm (0.238 in)
outer diameter and a 5.6 mm (0.218 in) inner diameter. A bill of materials is shown in Table 4-2.
This tube is inserted into one of the two, “half-inch” flanges. A UV detector is mounted in the
opposing 12.7 mm (0.5 in) port. After ignition, the ignitor is removed and replaced by a second
UV sensor. Each component in the ignitor has a high melting point that allows the ignitor to survive
for some time in the presence of a flame but is removed to avoid unnecessary exposure.
35
Figure 4-2. Image of the ignitor showing the spark between the inner tube and hook on the outer
tube.
A large voltage is supplied to the inner tube by a Ferromagnetic Ignition Transformer,
model 1092 type F, made by Allanson International, Inc. The single pole, one end ground
transformer has a maximum secondary voltage rating of 6000 VRMS and a secondary short circuit
current rating of 20 to 25 mA. The transformer is powered by a primary operating input of 110
VAC. The transformers loading air gap is 1.59 mm (1/16 in) to 3.18 mm (1/8 in). The inner tube
is well insulated from the outer tube by the center tube except for the tip where a small hook is
welded onto the outer tube and reaches within approximately 2 mm from the inner tube. The outer
tube is grounded to the reactor when placed in the “half-inch” port. This allows a spark to form
between two conductive tubes. The ignitor has been successfully used for perhaps twenty or more
start-up events to date and continues to produce a strong spark. Recently, there have been issues
36
with the alignment of the two tubes such that the ignitor is not seen by the UV sensor when both
are in their designed positions because the two holes were not drilled in perfect alignment. The
ignitor itself has not failed.
Table 4-2. Ignitor bill of materials.
Part Name Supplier Item Number 316 Stainless Steel Outer Tube Pipe Valve and Fitting Co. NA Alumina Tube JMS SOUTHEAST 5DAP14FZ 316 Stainless Steel Inner Tube McMASTER-CARR 89935K22
4.2 Pressure Certification
The POC size and pressure meets the requirements of a “pressure vessel” and must
therefore be certified to be safe to operate at the design pressure before its first operation. The
inspection process involved, among other things, demonstrating the presence of a rupture disk that
will depressurize the reactor if the pressure rises above 2.6 MPa (380 psig). While a disk was
present, the inspection resulted in certification only if the disk was changed to an ASME certified
disk. As a result, the existing disk was replaced. A test was also conducted by pressurizing the
POC until the rupture disk burst; this ensures all components can operate at a higher pressure than
the rupture disk. Prior to pressure testing, all flanges were tightened to the appropriate torque
settings for spiral wound gaskets. The pressure test was successful with the rupture disk failing
near 2.6 MPa (380 psig).
4.3 Software – OPTO 22
A Programmable Logic Controller (PLC) monitors the reactor during operation. The PLC
monitors the output of sensors, and various inputs and uses the measured signals to determine if
37
the reactor needs to shut down. With this capability, the reactor will be able to run without in-
person supervision, enabling overnight heat-ups. Sensors are mounted strategically throughout the
system to provide signals at critical locations. OPTO charts are the software modules where code
is written to produce output signals based on input sensors and control reactor operation. OPTO
GROOV is an application (AP) that can be installed on a cell phone or WIFI device that enables
remote monitoring of the system. The GROOV display can access the PLC from anywhere with
any device that can receive data. PAC User Display is the name of the software that is used in the
OPTO package for on-site monitoring. It has a GUI for an operator to easily monitor operating
conditions and control outputs on site.
4.3.1 OPTO Control Box
In the ERL 106C, an OPTO Control Box is found in the southeast corner. A layout of this
box is shown in Figure 4-3. The box contains four module mounting racks. A module mounting
rack has a SNAP-PAC-R1 controller (white) and modules for processing I/O points (blue). For
example, a module mounting rack has been dedicated for thermocouples. This module has an R1
SNAP-PAC-R1 controller (white, labeled TC Controller) and four thermocouple modules (single
blue box labeled terminal blocks) for connecting thermocouples to the controller. The controllers
send signals to the OPTO “Brain” through the network. The OPTO Brain is a piece of hardware
that processes signals from each SNAP-PAC controller and transmits them to the on-site computer
through the local network. Each controller has a dedicated power supply that delivers 24 volts to
its respective rack.
38
Figure 4-3. Diagram of the OPTO Control Box.
An ethernet port mounted in the OPTO box allows all four OPTO Controllers and the
OPTO Brain to communicate through the network. Each device is assigned a static IP address to
simplify communicating with OPTO. Additional power supplies provide twenty-four, twelve, and
five volts to POC hardware.
Black (ground) and red (hot) terminal blocks are VDC power supply outputs used for
powering hardware. Black (load), white (common), and green (ground) terminal blocks are 110
VAC power outputs for powering high voltage hardware. Since black terminal blocks were used
for VDC and VAC power outputs it is important to not use them interchangeably. For example, a
hardware’s VAC load wire should not be connected to a black VDC ground terminal block.
39
Since the control box is mounted in the same room as the reactor, a cyclone cooling
system was installed to prevent overheating. A solenoid allows the building’s compressed air to
flow into the cyclone which separates the air into cool and hot flows. The cool flow enters the
interior of the OPTO box. A fan is mounted to the bottom of the box to allow air to exit. The
solenoid is controlled by a thermocouple that is installed inside the box.
OPTO currently has one hundred and sixteen I/O points. All wires are carried through
conduit raceways to the OPTO control box. Appendix B. contains tables of the OPTO name,
location, and description of every I/O point currently in the OPTO Pac Controller.
4.3.2 OPTO PAC Control (Charts)
OPTO charts are visual flowcharts consisting of blocks that represent segments of code.
The chart shows the sequence blocks followed to complete certain tasks. Charts were written to
control important functions such as shutting down the reactor, adjusting the spray quench pump’s
rotational speed based on thermocouple readings, and calculating the refractory’s heat loss. Each
chart was written to perform one of these functions. A list of every chart and its function is
shown in APPENDIX E. OPTO PAC CONTROL (CHARTS). contains a detailed description of
each chart and its logic.
40
Table 4-3. The control system’s charts and the function of every chart.
Chart’s Name Chart’s Function Coal Mass Flow Calculates the coal mass flow rate from the load cell
measurements. Pressure Control Valve Controls the position of the Pressure Control Valve. Bypass Valve Interlock Software lock that doesn’t allow the Bypass Valve to
open if the reactor’s pressure is above a threshold. System Interlock Controls the two ball valves that remove the condensate
from the heat exchanger. Control Box Cooling Activates the control box cooling if the temperature
within the box is above a certain threshold. Emergency Trigger Logic Monitors critical sensors on the reactor to decide if the
reactor is operating safely or needs to shut down. Emergency Shutdown If the Emergency Trigger Logic Chart decides that the
reactor needs to shut down this chart this chart shuts down the reactor.
Heat Loss and Inside Temperature Calculates the heat loss and the refractories inner surface temperature from the B thermocouples measurements.
Ignitor Turns on the Honeywell box permissions and the ignitor. Quench Spray VFD Controls the spray pump’s rpms, and the amount of water
introduced upstream of the stainless steel connector, based on the measured stainless steel surface temperature.
4.3.3 OPTO PAC Display
The OPTO PAC Display provides a GUI for an operator to monitor and control I/O points.
The reactor’s system fills three monitors. An image of the middle monitor is shown in Figure 4-4.
The large window contains an image of the reactor system from the burner to the exhaust ducting.
Inputs and outputs are located on the schematic diagrams near their actual physical locations.
Inputs that can be changed by the operator are outlined in red whereas outputs are not outlined. On
the left side nine blue buttons are used for opening display windows. A text box displays the error
messages from the “Emergency_Trigger_Logic” chart. An image of the alarm lights displays the
status of the light stack above the ERL106C door. A green light communicates that the reactor is
operating under atmospheric conditions, a yellow light is for pressurized operation (the reactor is
41
above 69 kPa (10 psig), and a red light means the reactor shutdown chart was started. The two
buttons next to the light stack image display the statuses of the “Emergency_Trigger_Logic” and
“Emergency_Shutdown_Procedure” charts and will change the status of the chart if clicked.
The top right window displays every B thermocouple signal and magnet K thermocouple
temperature. A window located in the bottom right displays the outputs and inputs associated with
the OPTO control box.
Figure 4-4. Middle monitor of the OPTO display.
The left monitor is shown in Figure 4-5. Five windows are shown on this monitor. The
leftmost window displays the coal feed hopper. Inputs and outputs are located where they are
physically located on the hopper. The coal mass flow rate was designed to be controlled by
adjusting the CO2 fluidization flow rate and vent flow rate. The mass flow controller that
introduces the fluidization CO2 can be adjusted manually from the display. The vent flow is
42
controlled by a rotameter and must be changed by hand. An OPTO PID was also created for the
CO2 fluidization to be adjusted automatically based on the calculated coal mass flow rate. A box
in the middle of the window contains controls associated with the PID. To this point, the
fluidization CO2 mass flow rates have only been adjusted manually.
The condensation control window is in the top middle of the monitor. This window displays
all the controls associated with the condensation ball valves. Indicators display if the top and
bottom ball valve positions are closed or open. The condensation delay may be set in the text box.
This delay controls the time between the end of one cycling of the ball valves until the ball valves
start cycling again.
The chart window is in the top right of the monitor and displays the status of each chart. A
green button indicates that the chart is currently running and by clicking the green button the chart
will stop running. A red button indicates that the chart is stopped and by clicking a red button the
associated chart will start running. The “Data Collection” box does not control a chart, but
indicates data is being recorded. If the data collection is turned on, predetermined inputs, outputs,
or variables are saved to the public folder. Because they are saved to the public folder anyone can
access them from the lab computer. Currently, the historical data logs are programmed to update
every minute, except for the coal feeder I/O points update every five seconds.
The bottom right window has all the controls associated with the burner. The seven mass
flow controllers that introduce gas to the burner can all be set and monitored in this window. The
solenoids upstream of each mass flow controller are displayed and can be opened or closed by
clicking on the button. Two additional solenoids that are located on the high pressure air lines are
shown in the bottom left of the window. On the bottom right, a box contains three boxes that are
associated with the ignitor chart. The left button is the ignitor chart status, the middle button is the
43
Honeywell box permissions, and the right button is the ignitor. If the ignitor chart runs successfully
and the high pressure natural gas solenoid is opened by the Honeywell box, then a flame symbol
is displayed on the right side of the window. Above the flame, an ignitor image is visible or
invisible indicating when the ignitor is on or off.
The reactor control valve chart is located in the middle of the window. An image of the
three valves are displayed with controls located next to their respective valve. The compressor air
pressure is displayed in this chart because two of these valves are pneumatic and are controlled by
the high pressure air. A box containing the controls associated with the Badger valve is displayed
in the bottom right of this window. The bottom middle window contains all the exhaust gas
concentrations and a button to turn the zirconium oxygen sensor on and off.
Figure 4-5. Left monitor of the OPTO display.
The right monitor contains the plot window. Six plots are displayed in this window. Each
plot is a real time I/O value from the reactor. From the top left and moving clockwise these are
44
refractory and flue gas temperatures, hopper coal mass, coal mass flow rate, reactor pressures,
Zirconium oxygen concentration and Horiba gas concentrations. Each plot is currently configured
to update every minute.
4.3.4 OPTO GROOV BOX
The GROOV Box provides a GUI for users that can be viewed remotely. The interface is
customizable for computers and mobile devices. This allows any device connected to BYU’s
virtual private network (VPN) to monitor the reactor remotely. This is important for an overnight
heat up before introducing coal the following morning. It is accessed through any browser. While
on BYU’s local network the following is entered http://10.8.113.40. This IP address is dedicated
to the GROOV Box’s Mac address and will not change.
4.3.5 POC Camera
An internet protocol camera from Amcrest, model number IP2M-841, is also used for
monitoring the reactor remotely. It produces a real time image of the flame seen through the
sapphire window located closest to the burner. Like the GROOV Box, the camera may be accessed
by any device on BYU’s local network. To access this camera the IP address is entered into any
browser window. Unlike the GROOV Box, the camera does not have a dedicated IP address and
must be periodically checked using an AMCREST program that displays the IP address of
AMCREST cameras on your network.
4.4 Standard Operating Procedure (SOP)
A Standard Operating Procedure (SOP) was written to ensure the safe operation of the
POC. It provides step by step instructions to the startup, operation, and shutdown of the POC
45
system. It can be printed out and stepped through as a checklist while operating the POC. This
document is found in APPENDIX B. STANDARD OPERATING PROCEDURE
4.4.1 Regular Maintenance List
In addition to the SOP, a document with a list of all the items that require regular
maintenance has been produced. It includes the name of the part, what maintenance is required,
and how often it must be done. This document is found in APPENDIX C. REGULAR
MAINTENANCE LIST
4.4.2 Fuel Air Ratio Calculations
An Excel sheet was created for future operators to know what flow rates are needed to
achieve safe fuel-to-air ratios. For each firing rate a natural gas flow rate is specified with the air
flow rate for three percent oxygen in the exhaust. Three percent oxygen in the exhaust is standard
in industry to ensure all hydrocarbons have sufficient oxygen to combust.
In addition to air flow rates, a sheet was made for pseudo air operation. This sheet allows
the operator to specify the mass fractions of oxygen to carbon dioxide along with the percentage
of gas introduced to both locations of the burner. Based on this information, the mass flow rate for
both oxygen and both carbon dioxide mass flow controllers are calculated. This allows the user to
know what flow rates to operate at for any firing rate or natural gas flow rate.
46
The entire POC system has been operated with atmospheric natural gas, pressurized
natural gas, and pressurized coal as a path toward the final objective of obtaining temperature,
heat flux and radiation data under pressurized coal combustion. This chapter presents the results
of these total system tests leading to and culminating in a final test that obtained pressurized dry
feed coal results with radiative and wall temperature heat flux measurements. This chapter will
describe the functionality of all of the subsystems, issues that were identified and corrected and
issues that remain to be resolved. It will also describe the flame shape based on heat flux
produced by a single set of fuel, carrier gas and oxidizer flow rates.
5.1 Atmospheric Natural Gas Operation
An atmospheric natural gas flame was required in order to cure the refractory and to heat
the reactor walls in preparation of for coal combustion. After pouring, the refractory must be cured
to remove moisture and set the material. Also, before introducing coal, the refractory wall must be
heated above the auto ignition temperature of the coal, which was determined to be above 816 oC
(1,500 oF). Ideally, the reactor’s walls are heated above 1316 oC (2,400 oF), the slagging
temperature, before starting coal combustion. A natural gas flame is used to preheat the refractory
to these temperatures before coal is introduced. The combined fuel, burner, reactor, and exhaust
5 RESULTS
47
sub-systems were tested under atmospheric conditions to determine if they could achieve these
objectives.
5.1.1 Refractory Curing
The dryout and curing schedule for the refractory as suggested by the manufacturer is
shown in Table 5-1. The system was limited to heating from the combustion products produced
by the burner. High pressure (3.1 MPa) natural gas was available from the nearby campus heat and
power station and was selected as a fuel source. Heat from a flame results in a non-uniform axial
temperature profile and little control in holding the temperature of the refractory uniform or
constant.
Table 5-1. Curing and Dryout Schedule for UltraGreen SR (Castable) Refractory.
Curing and Dryout Schedule (heated from the hot face) Hours Cure Time
Minimum Ambient Cure Time 24 Dryout Heat/Hold Sequence
Heat from ambient to 150oC at 28oC/hr ~5 Hold at 150oC for 1 hr for each 25.4 mm. (1 in.) of thickness up to 304.8 mm. (12 in.)
4
Heat from 150oC to 371oC at 28oC/hr 8 Hold at 371oC for 1 hr for each 25.4 mm. (1 in.) of thickness up to 304.8 mm. (12 in.)
4
Heat from 371oC to 649oC at 28oC/hr 10 Hold at 649oC for 1 hr for each 25.4 mm. (1 in.) of thickness up to 304.8 mm. (12 in.)
4
Total Dryout Time ~35 Total Curing and Dryout Time ~59
The refractory material had been set but not cured for over twelve months. It was therefore
assumed that significant drying had been achieved and that the initial low temperature heating
could be skipped. The curing process was then accomplished through two different heating
48
periods. The first heat-up was at the lowest possible firing rate in the reactor which was limited by
the UV detectors ability to see the flame as it became shorter than the viewing location. The level
3, outer, B-Thermocouple was monitored until the refractory reached approximately 371oC
(700oF). At this point, the heat up stopped, and the refractory surface energy was allowed to diffuse
through the refractory creating a relatively uniform temperature until it slowly cooled to the
ambient temperature over a period of several hours. For the second heat-up process, the refractory
was heated until the level 3, outer, B thermocouple read 649oC (1200oF). Then the heat up stopped,
and the heat diffused through the refractory slowly cooling down.
5.1.2 Reactor Preheating for Coal
A diagram of the burner’s ports is shown in Figure 5-1. Air is introduced using a single
mass flow controller to both the outer annulus and the eight tertiary ports. In this configuration the
distribution of air between the outer annulus and tertiary tubes is not controlled or known. It was
anticipated that at the designed firing rate of 100 kW a stable atmospheric natural gas flame would
be required to burn overnight to reach slagging wall temperatures. During this heat-up period, the
reactor quench spray would need to run constantly to prevent overheating of the exhaust pipe and
spray nozzle. The Bypass Ball Valve is left fully open to enable the reactor to remain at
atmospheric pressure. Purge gas is introduced through the thirteen designated port locations to
keep water vapor from condensing on the “two-inch” viewing port windows which are used for
the radiometers, the FTIR, two-color extinction measurements or a viewing camera. The full
procedure for preheating the reactor is contained in the SOP.
49
Figure 5-1. Burner Diagram with Dimensions [8].
5.1.3 Atmospheric Natural Gas Results
A stable flame has been achieved in the reactor with the burner operating between 25 to
115 kW using flow rates as outlined in Table 5-2 . The reactor has also been operated overnight at
firing rates between 70.5 to 100.5 kW. While the burner may also operate steadily at other air fuel
ratios, these were selected to match typical industry air-fuel ratios with 3% excess oxygen
assuming complete combustion.
50
Table 5-2. Atmospheric natural gas flow rates with respective firing rates
Natural gas (kg/hr) Air (kg/hr) Firing Rate (kW) 2.0 40.7 26.1 3.0 61.1 39.2 4.0 81.5 52.2 5.0 101.9 65.3 5.4 110.0 70.5 5.6 114.0 73.1 6.1 124.3 79.6 6.9 140.6 90.9 7.6 154.8 99.2 8.8 179.5 115.0
The reactor was designed to use gravity to remove water from the exhaust system through
the condensation ball valves located at the bottom of the heat exchanger. At firing rates exceeding
80 kW, it has been observed that the total gas flow rate in the exhaust exit causes water to be
entrained in the exhaust gas and carried through the heat exchanger, accumulating in the cyclone
barrel. Along this path, the water is transported past the zirconium oxygen sensor impinging and
causing the sensor failure. An “L” extension was used to move the sensor from the direct flow of
exhaust with a smaller slip flow bringing a portion of the exhaust gas past the oxygen sensor. These
changes reduced the occurrence of oxygen sensor failure but did not eliminate it. To avoid water
damage, the reactor must operate at or below a maximum firing rate of 80 kW with natural gas
combustion. This limit does not apply to oxy-combustion because less total gas is required to burn
the fuel and velocities are reduced.
The data for the plots shown in this section were taken on April 29, 2021, to April 30, 2021.
Before starting data collection on April 29, the refractory was already 500oF. The reactor was
operating at a firing rate of 79.6 kW. A plot of the refractory surface temperature and heat loss
calculated from the level 1 and 5-inner and outer B-thermocouple pairs are shown in Figure 5-2
51
and Figure 5-3, respectively. The shape of the curve representing the surface wall temperatures
can be seen to rise rapidly and then begin to asymptote to a final temperature. The driving force of
heat transfer is the temperature gradient. The temperature gradient between the refractory and the
hot gas was large initially; this resulted in the refractory temperature increasing quickly at first. As
the refractory approached the gas temperature the heat transfer slowed resulting in a much slower
increase in temperature.
Figure 5-2. Refractory inner wall temperature calculated from the B-Thermocouples during heat-
up at a firing rate of 79.6 kW.
Although the refractory had not reached a steady-state temperature at 9:05 am, after 20+
hours of heat-up, the top level (level 1) nearest the burner reached the temperature required to
ignite coal (816oC) in less than two hours and the bottom of the reactor (level 5) in less than four
hours. As will be discussed later, the natural gas flame would not operate under pressurized
conditions and the coal was not designed to be fed at atmospheric conditions. Therefore, there is
52
a need to overheat the reactor because the surface temperatures will cool rapidly during
pressurization. Heating the reactor on natural gas and air also saves money as they are less
expensive than compressed O2, CO2, and pulverized coal.
Initially, the refractory at the bottom of the reactor was at a lower temperature than the
top of the refractory. However, after eight hours of heat-up the peak temperature shifted to the
bottom of the reactor. This can be explained by a premixed rich flame near the top of the reactor
created by the air in the outer annulus reacting with the natural gas in the center tube followed by
a diffusion or mixing limited flame further down the reactor created by the eventual mixing of
the tertiary air into the fuel rich products of the premixed flame. Initially heat is removed from
the premixed flame near the top of the reactor heating the top section rapidly and reducing the
temperature and energy remaining in the diffusion flame. As the upper wall temperatures rise
and remove less heat from the premixed flame, the temperature of the diffusion flame increases,
and the walls eventually reflect the temperature profile of the flame in a near adiabatic state
where the peak temperature is achieved at the point of complete combustion or the end of the
diffusion flame.
The heat loss through the refractory calculated from embedded thermocouples is shown in
Figure 5-3. The shape of the heat loss curve vs. time is seen to increase from zero to a maximum
and then decrease from the maximum to a steady-state asymptote. The heat loss starts at zero
because the temperature on the inside refractory surface and outer shell are the same and there is
no temperature gradient. When combustion occurs, the temperature of the inner shell rises rapidly
increasing the temperature gradient between the inner and out surfaces and creating a large
temperature gradient. Over time, the inner temperature remains relatively constant while the outer
shell temperature increases causing a decrease in temperature gradient and a decrease in heat loss.
53
Eventually, the inner refractory temperature, the outer shell temperature and the temperature
gradient reach a steady value and heat loss becomes steady. It should be noted that this heat loss
only represents the loss of heat from the gas to the inner wall when the reactor is at steady-state.
Assuming an average heat loss for the whole reactor as the average of these two heat losses,
the total heat loss for the reactor would be six times the average heat loss. The maximum heat loss
would be at approximately 16.8 kW at the start and slow to approximately 6.6 kW. This would be
21% of the energy being removed from the flame at the start of the heat-up period and 8.2 % of
the energy being removed at the end of heat-up. The gas temperature at level 5 of the reactor was
not measured, but assuming the wall temperature will asymptote to the gas temperature, it appears
that the gas temperature is somewhere near 1800 K. An energy balance for methane combustion
at the experimental fuel air conditions suggest 18% of the heat must be removed to reach this
temperature. This suggest that the heat loss values are reasonable but perhaps a little lower than
expected.
Figure 5-3. Refractory heat loss calculated from the B-Thermocouples during heat-up at a firing
loss of 79.6 kW.
54
The maximum allowable reactor shell temperature is 232oC (450oF). Eight surface
thermocouples positioned at various locations on the reactor monitor the shell temperature. A plot
of the three hottest measured surface thermocouples temperatures is shown in Figure 5-4 . The
measured reactors shell temperatures are well below the maximum allowable temperature at the
eight measured locations. The shell surface thermocouples increased linearly for the first seventeen
hours and then plateaued for the last couple hours of heat-up.
Figure 5-4. Surface magnet thermocouples placed on the reactor’s shell.
The Horiba analyzer was used to measured O2, CO2, CO, SOx, and NOx concentrations.
The oxygen and carbon dioxide measurements during the heat-up are shown in Figure 5-5 . The
Horiba was not calibrated on April 29, the day these data were taken. The oxygen fluctuates around
4 % by volume, which is slightly higher than the 3 % value expected from the fuel and air flow
rates. The Horiba samples are taken from the level 5 southeast access port where it was assumed
55
combustion would be complete. The natural gas and air flow rates are 6.1 kg/hr and 124.3 kg/hr,
respectively. The initial readings of 21% O2 and 1% CO2 are a measure of air before a flame is in
the reactor. This clearly shows the CO2 was not calibrated. The constant concentration
measurements over the nineteen-hour heat-up indicate the stable atmospheric natural gas operating
conditions of the reactor.
Figure 5-5. Horiba oxygen and carbon dioxide concentration measurements during heat-up.
Temperature measurements are shown in Figure 5-6 for three separate locations in the
exhaust: the reactor exit, the entrance to the heat exchanger and the exit of the heat exchanger. The
exit temperature is seen to rise rapidly initially and more slowly as the reactor approaches a steady
temperature. At the end of the heat-up, the temperature is 1528 K. The energy in the gas at this
point is approximately 68% of the heating value of the fuel. The gas temperature is approximately
300 K below the wall temperature at level 5 indicating significant heat loss between level 5 and
the outlet.
56
The heat exchanger inlet temperature varies significantly over a range of approximately
100 oC. This is caused by the water vapor spray that is introduced after the Flue Gas exit
thermocouple to keep the stainless-steel pipe between the reactor outlet and the heat exchanger
cool.
The heat exchanger inlet temperature is also the exit temperature of the 310 stainless-steel
connector and is seen to be over 800oC which is well above the maximum allowable surface
temperature of 503 oC. Liquid water from the spray quench and an external exhaust fan keep the
surface temperature of the pipe below that of the gas as will be seen in the next section. The heat
exchanger outlet temperature is consistently 26.7 oC, indicating the heat exchanger is effective at
removing heat from the exhaust gas.
Figure 5-6. Flue gas measurements from probe K-Thermocouple in the flue gas.
57
Temperature measurements of the surface thermocouples (K-type) that monitor the
temperature of the inlet and outlet of the stainless-steel connector and the weld for the connection
between the stainless-steel connector and heat exchanger are shown in Figure 5-7. The stainless-
steel surface thermocouples are at least 50 oC below the maximum allowable temperature of 503
oC. Recently, a fan has been used to cool the stainless-steel connector in addition to the spray
quench. For this heat-up, a fan was not used. These data were obtained with the spray nozzle from
Spray Systems Co. installed onto the Quench Spray Subsystem.
Figure 5-7. Surface measurements from surface K-Thermocouple.
A plot of the temperature measurements of the coolant entering and leaving the heat
exchanger is shown in Figure 5-8. The coolant system ball valves were fully open allowing the
maximum flow of coolant to the heat exchanger. The coolant flowmeter measured between 43.5
and 44.5 LPM consistently during the heat-up process. The coolant inlet temperature was slightly
58
below 16 oC, and the coolant out temperature increased from 27 oC to 38 oC during the duration of
the heat-up. The increase in the temperature of the coolant leaving the system is due to less heat
being lost through the refractory, as the refractory heats up. Assuming an outlet coolant
temperature of 38 oC, inlet temperature of 13.6 oC, and a flow rate of 44 LPM the calculated heat
transferred to the coolant in the heat exchanger is 73.8 kW. This is 93% of the heat available for
complete combustion of the fuel. This value seems high but is consistent with estimates from the
heat loss measurements discussed above which suggest approximately 8% of the heat is lost in the
refractory lined section of the reactor.
Figure 5-8. Coolant temperature measurements from probe K-Thermocouples.
5.2 Pressurized Natural Gas
The reactor’s coal feeder was designed to only function at high pressures. For this reason,
the reactor must be pressurized before the feeder can introduce coal. One method of pressurizing
the reactor without cooling the refractory is high pressure natural gas. Ideally, a natural gas flame
59
would continue to burn during the pressurization process. This section discusses attempts at
operating with a high-pressure natural gas flame in the POC.
5.2.1 Method
As demonstrated in Section 5.2, the reactor can be heated with an atmospheric natural gas
flame at 80 kW. Before switching to pulverized coal, it is desirable to pressurize the reactor with
a continuous flame in order to maintain high wall temperatures. There are two valves typically
used to control the pressure within the reactor: a “two-inch” electrically driven ball valve and a
“half-inch” pneumatic valve (Badger Valve). The ball valve can be either completely open or
closed. In the open position, the reactor will not pressurize. A third valve, the emergency bleed-
off or drain valve can also be opened or closed. It allows reactor gas to pass through a 3.56 mm
orifice. Opening this valve will also decrease the rate at which pressure rises in the reactor but will
not prevent the pressure from rising when air or CO2 are flowing into the reactor at typical
operating flow rates. The Badger valve may be controlled to be open from 0 to 100 percent and is
used as a fine control of the reactor’s pressure. When the Badger valve is in the 100% open
position, the opening is small enough (12.7 mm) that the reactor will begin to pressurize.
Therefore, there is no process available to hold the reactor pressure steady in the range of
atmospheric pressure to 276 kPa or slightly above atmospheric pressure.
The standard atmospheric operating procedure is to introduce natural gas through the inner
annulus and air through the burner’s outer annulus and eight tertiary tubes. It would therefore be
best if flow through these same locations were maintained when switching to pressurized
combustion. The distribution of air between the outer annulus and tertiary tubes is not known and
cannot be controlled with a single air mass flow controller. Another issue also discovered after
testing began is that the air mass flow controller was not designed to operate above atmospheric
60
pressure. The spring in the mass flow controller that controls the air mass flow rate was not selected
for pressurized operation.
5.2.2 Pressurized Natural Gas Results
Upon closing the Bypass Valve, and with the Badger valve completely open, the reactor
pressure would begin to increase as expected. Visual observations of the flame showed the color
of the flame to shift from a continuous blue to a flickering yellow. Somewhere before reaching
293 kPa, the UV detectors would shut off the fuel flow and the flame would go out. It was unclear
if the flame shortened and was no longer in the view of the detectors or if the flame went out
causing the detectors, to shut off the fuel. This same procedure was repeated multiple times with
the same result.
The change of flame color suggests the flame is transitioning from a premixed flame that
is stabilized by the swirled flow in the outer annulus to a non-premixed diffusion flame. As the
pressure of the flows approximately doubled, the velocity of both the fuel and oxidizer should have
decreased by a factor of two. An attempt was made to determine if the flow through the annulus
may have also shifted from the annulus to the tertiary tubes which could also reduce mixing.
A second mass flow controller was added to the air line to control flow independently to
the annulus and tertiary lines. This mass flow controller was calibrated for use with CO2 which
made the flow rate difficult to quantify precisely, but it did allow independent control. Different
distributions of air to each port were attempted, but when the pressure was increased, none were
successful at producing a flame that could be seen by the detector.
Next, two CO2 and O2 mass flow controllers were used to supply a mixture of CO2 and O2
independently to the outer annulus and tertiary lines. The following variables were changed
systematically in attempts of achieving a pressurized natural gas flame: O2 to CO2 mass fraction,
61
the port where natural gas was introduced (inner annulus and combined center tube and inner
annulus), firing rate (methane flow rate), fraction of CO2 / O2 mixture introduced to each port, and
the position (open or closed) of the emergency drain valve. Table 5-3 shows all the combinations
of variables that were investigated. The purpose of opening the emergency drain valve was to
decrease the rate of pressure rise in the reactor. The highest pressure that was achieved with an
operating flame was still 207 KPa.
Table 5-3. Different operating conditions tried to achieve a pressurized natural gas flame.
Mass fraction (%)
Natural Gas Port Firing Rate (kW)
Inner Annulus (%)
Emergency Drain Status
29 Inner annulus 25 75 Closed 29 Inner annulus 40 75 Closed 29 Inner annulus 60 75 Closed 29 Inner annulus 40 100 Closed 29 Inner annulus 40 75 Closed 29 Inner annulus 40 50 Closed 29 Inner annulus 40 25 Closed 29 Inner annulus 40 0 Closed 35 Inner annulus 40 75 Closed 35 Inner annulus 25 75 Closed 35 Inner annulus and primary tube 40 75 Closed 35 Inner annulus and primary tube 40 75 Closed 35 Inner annulus and primary tube 40 75 Open 35 Inner annulus and primary tube 25 75 Open 35 Inner annulus and primary tube 25 25 Open
It was concluded that regardless of the amount of flow through the outer annulus, the
lower velocities produced by pressurizing the reactor caused a change in the flow dynamics of
the swirled flame such that the flame transitioned to an unsteady diffusion flame.
Although desirable, high pressure gas combustion is not essential for operating a coal
flame. While pressurized natural gas combustion would allow a transition to coal combustion
62
without losing heat from the walls, the thermal capacitance of the reactor walls is large enough to
allow coal to ignite after the reactor is pressurized with inert gases. It was therefore decided not
to pursue additional testing of a pressurized natural gas flame. One obvious choice for additional
study would be a premixed natural gas flame where oxidizer is added to the fuel upstream of the
exit. Premixing has not been attempted because of the potential for flashback. Flashback is when
the flame propagates upstream into the burner because the flame speed is higher than the
premixed gas velocity at the burner tip. Maintaining a gas velocity higher than the flame speed
would be hard to accomplish because while pressurizing the reactor from atmospheric to 20 bar
would result in the gas velocity decreasing by a factor of 20.
5.3 Pressurized Coal
This section summarizes testing done with pressurized coal with the goal of producing a
steady-state coal feed and combustion measurements.
5.3.1 Method
The reactor was first heated up the day before testing with an atmospheric natural gas flame
at a firing rate between 70 to 80 kW. The natural gas flame was left on overnight and the reactor
diagnostics were monitored remotely using the GROOV AP. The following morning the coal
feeder was filled with coal to avoid compaction or the possibility of absorbing moisture overnight.
After reaching or approaching the desired slagging wall temperatures of 1589 K (2400 oF)
the natural gas was turned off and CO2 was used to pressurize the reactor. When the reactor and
the coal hopper were close to operating pressure (approximately 1.4 MPa), the flow rates to various
locations were set as shown Table 5-4. The coal feeder flow rates were chosen to match those used
by Tuia [7] for 100 kW. The burner flow rates were chosen to match those recommended by
63
Carpenter [8]. A summary of the target nominal flow rates are found in Table 5-4. These flow rates
were found from modeling results to produce temperatures high enough to produce slagging walls
but low enough to protect the refractory (2144 K, 3400 oF max) and the outer shell (505 K, 450
oF). As the cold CO2 entered, the reactor wall temperatures dropped, but if they remained above
1000 K (1340 oF) it was assumed to be safe to introduce coal.
Table 5-4. POC target mass flow rates for pressurized coal operation.
Fluid / Material System Location Flow Rate (kg/hr) Air Burner Outer Annulus 0 Low Press. NG Burner Center Tube 0 O2 Burner Tertiary Tubes 26.9 O2 Burner Inner Annulus 3.2 CO2 Burner Tertiary Tubes 26.9 CO2 Burner Inner Annulus 12.6 High Press. NG Burner Center Tube 0 CO2 Coal Feed Dilution CO2 9.7 Fluidization CO2 Coal Feed Tuyere Fluidization 4 Coal Coal Feed Hopper Exit 12.3 CO2 Coal Feed Hopper Vent 1.5
In order to understand the results it is important to briefly review the principles behind the
fluid bed coal feeder as presented by Tuia [7] . The coal feed system shown schematically in Figure
5-9 has two controlled mass flows labeled ��𝑚𝑓𝑓 for fluidization CO2 and ��𝑚𝑑𝑑 for dilution CO2. A
third flow, ��𝑚𝑣𝑣, the vent CO2 flow, exits the top of the coal hopper and is measured by a rotameter
and is controlled manually. The fluidization flow fluidizes the coal at the base of the hopper and
then carries pulverized coal particles toward the reactor. Tuia [7] showed that the mass flow rate
of coal conveyed from the hopper to the reactor is proportional to the difference in the fluidization
and vent flow rates of CO2 or the net flow of CO2 from these two flows. The dilution flow helps
64
to carry CO2 to the reactor but does not entrain additional coal beyond what is entrained by the
fluidization flow and does not impact the coal flow rate.
Figure 5-9. Diagram of coal feed flows
For the tests presented here, the nominal operating pressure for the reactor was set to 1.46
MPa (200 psig). Prior to the introduction of coal, the coal hopper must be at a slightly higher
pressure than the reactor to avoid hot gases flowing into the coal hopper when the solenoid valve
connecting the two chambers is opened. The initial pressure difference between the two chambers
was set at 137.4 kPa (20 psi) but was altered as will be described in the results to a current value
of 27.5 kPa (4 psi). When both the absolute pressures and the pressure differential were at desired
values, and all flow rates set as desired and described in Table 5-4, the valve between the two
chambers was opened enabling coal to be introduced to the burner. The O2 and CO2 were
introduced through the burner’s inner annulus and tertiary tubes as described in Table 5-4. The
coal, carried by CO2, was introduced through the center primary tube.
65
5.3.2 Pressurized Coal Results
Prior to writing this document there have been six attempts to introduce coal into the
reactor. The reactor is yet to demonstrate long term, steady, controlled combustion of coal. This
section will summarize results from these tests with three objectives; 1) to analyze as best as
possible the characteristics of the coal feeder using the limited load cell, radiometer, and
temperature measurements obtained to date 2) to analyze as best as possible the characteristics of
the coal flame within the reactor using the limited temperature, heat loss, and gas concentration
measurements obtained to date 3) to analyzer the ash management using images obtained after the
sixth attempt to burn coal. In depth details of every attempt to burn coal is shown in APPENDIX
G. PRESSURIZED COAL TESTS
It should be noted that after the fifth attempt to burn coal on May 28, it was discovered that
the coal feeder carbon dioxide mass flow controllers were scaled incorrectly. The scaling error
meant that the operator, for the first five attempts to burn coal, introduced 58 percent less carbon
dioxide in the fluidization and dilution flows than was previously believed at the time the data
were taken. All CO2 flow rates presented in this chapter are the actual flow rates that occurred in
the test, not the flow rates that were initially targeted unless the flow is identified as a targeted
value.
5.3.3 Coal Flame Analysis
A plot showing the refractory inner wall temperature measured during oxy-coal operation
Test #6 is shown in Figure 5-10. During oxy-coal operation the refractory is hottest at the top of
the reactor. This is evidence that the oxy-coal flame is short and most of the energy release is near
the top of the reactor.
66
It is important to note that the reactor has not reached steady state during oxy-coal operation
although it does appear that the inner wall temperatures are starting to asymptote at 1600oC. The
desired refractory temperature is between the slagging temperature (1300oC) and the maximum
refractory temperature (1900oC). This is evidence that the flow rates shown in Table 5-4 can
produce the desired refractory wall temperatures.
Figure 5-10. Reactor refractory wall temperature at two axial locations.
The calculated refractory heat loss during pressurized oxy-coal operation is shown in
Figure 5-11. Heat loss is significantly higher near the top of the burner than the bottom. This is
consistent with the higher measured temperatures at the top of the reactor during oxy-coal
operation and again is evidence of a short flame during oxy-coal operation.
The coal and natural gas heat losses are assumed to have similar shapes. Therefore, the
coal heat loss should decrease after reaching its maximum before it reaches steady state. Since the
67
coal has not approached steady state, conclusions cannot be drawn about the coal heat loss.
Although the coal heat loss has not reached equilibrium, the data show that the peak temperature
gradient is higher in this coal test than in the natural gas tested even though the outer shell
temperature and inner refractory temperatures were higher at the start of the coal burning test than
they were at the start of the natural gas test. This indicates that heat loss was higher for the coal
test even though the magnitude cannot be quantified. This suggests that the average heat loss for
pressurized oxy-coal operation will be higher than the measured average atmospheric natural gas
heat loss. While the coal flame was clearly visibly radiating more than the natural gas flame and
is expected to have a higher radiative component, it should be noted that the coal flow rate was
supplying 120 kWth in comparison to 80 kWth for the atmospheric natural gas firing rate.
Figure 5-11. Reactor refractory heat loss at two axial locations.
68
This analysis of the burner shows the flame characteristics can be tested with this system
during pressurized oxy-coal operation. In particular, heat loss, refractory wall temperature, and
gas concentration measurements give insight into burner performance. This system is capable of
being used to develop and diagnose future burners for pressurized oxy-coal conditions.
5.3.4 Coal Feeder Analysis
A plot of the radiometer and coal mass flow measurements recorded during pressurized
oxy-coal Test #6 operation is shown in Figure 5-12. In this plot the solid lines are averaged over
two minutes and the error bars represent one standard deviation. Although the coal mass flow rate,
which is calculated from load cell measurements, continually has a large standard deviation the
averaged mass flow rate is relatively constant. This is evidence that there are fluctuations in the
feeding system occurring in periods on minute or less.
Figure 5-12. Coal mass flow rate and three radiometer signals.
69
Temperature, heat flux and radiometer measurements indicate a flame is present when the
coal mass flow rate starts increasing at 9:00. Initially the coal was exceeded the target of 15.0 kg/hr
producing fuel rich conditions for the first five minutes. After the first five minute, the coal mass
flow rate decreased, and the reactor operated lean until coal quit flowing at 9:50. When the coal
quit flowing it was revived by vibrating the reactor. This is evidence that the coal is bridging. After
the coal flow rate was revived it fed at a slightly higher flow rate than before and it operated for
another twenty minutes until the coal feeding was stopped.
Three radiometer signals are also shown in Figure 5-12. The signals are reported as
voltages because the radiometers have not been calibrated prior to this text. The radiometer signals
fluctuates when coal is burning. Initially when coal was first introduced, and the reactor was
operating rich, all three radiometers show fluctuations indicating a flame. After several minutes
when the coal mass flow rate decreased, and the reactor was operating lean the Level 2 and Level
5 radiometers quit fluctuating. The indications of lower-level radiometers fluctuating when fuel
rich but not when fuel lean is evidence that a short flame is produced unless there is too much coal
for the oxygen being introduced, then coal burns throughout the entire combustion chamber.
A plot zoomed in to show a higher resolution one minute section of two of the radiometer
signals is shown in Figure 5-13. The sampling rate obtained was every second although a higher
sampling rate is possible. Fluctuations in Level 1 are high while barely measurable in Level 2.
This is evidence that during stoichiometric operation, radiative emission from combustion is
primarily limited to the firs level of the reactor.
70
Figure 5-13. One minute of radiometer signals.
Although the load cells produce high frequency noise, they do not have the resolution to
show high frequency fluctuations in the coal feed. However, the radiometers are capable of
diagnose higher frequency fluctuations due to their fast sampling rate. In Figure 5-13, both
radiometer signals are fluctuating every three to four seconds. These fluctuations are evidence of
intermittent puffs of coal producing highly radiative intermittent combustion events. This matches
visual observations through the top viewing port. When viewing through the port, a burst of light
was observed every few seconds followed by no light. This cycle repeated continually whenever
feeding coal into the reactor.
This analysis shows that the coal feeder can be tested during pressurized oxy-coal
operation. The radiometers can provide high frequency information while the load cells produce
71
low frequency averaged coal mass flow rates. Several problems have already been diagnosed with
this system such as bridging and plug flow.
5.3.5 Ash Management Results
After the reactor had operated for several hours of oxy-coal combustion the ash tank of the
reactor was removed, and ash deposition was checked. Two images of the ash that accumulated in
the ash tank are shown in Figure 5-14. The picture on the left is of the entire ash tank and the
picture on the right is zoomed in to show the ash that had accumulated in the tank. From the
pictures it is apparent that slag dropped onto the refractory as designed.
Figure 5-14. (left) zoomed out picture of the ash tank outside the reactor. (right) zoomed in
picture of the ash that has built up in the ash tank.
The ash tank is about a third full after several hours of operation. The amount of ash that
this tank can hold may be the limiting factor for how long the reactor can burn coal continuously.
72
An image looking up into the reactor from the bottom after hours of burning coal is shown
in Figure 5-15. Black lines are from molten slag flowing down the reactor walls. Even though the
reactor had periods of very coal rich operation prior to this picture being taken, subsequent periods
of oxidation following the coal feed period allowed the carbon to burn off the walls and minimal
ash deposition is seen. This is likely due to the coal depositing initially on the walls and once the
coal oxidizes the ash it produces eventually slags running down the walls into the ash tank as
designed.
Figure 5-15. An image looking up into the combustion chamber of the reactor after several hours
of burning coal.
These pictures provide evidence that the ash management system of the reactor is
functioning as designed. The picture of the combustion chamber is evidence that the reactor will
73
likely not have significant ash deposition. Ash deposition will likely not be a limiting factor for
how long this reactor can burn coal continuously.
74
BYU has been developing, constructing, and testing a pressurized oxy-coal reactor for the
past five years. The purpose of this reactor is for testing a novel method for dry coal feeding at
pressure, a high pressure oxy-coal burner, and ash management. Prior to my work a burner, a
combustion pressure vessel, an exhaust system, and a preliminary coal feed system were designed,
constructed, and independently tested but never tested as a combined system. The objective of the
work reported here was to combine these individual components into a working system and
demonstrate the system’s ability to measure and develop critical high pressure coal combustion
technologies. Contributions to the overall project contributed by this work include:
• Completion of several reactor subsystem including 1) a swirler for the burner air 2) a
nitrogen purge system to keep condensate from forming on windows and 3) high pressure
windows and purge system for optical access.
• Design, fabrications and testing of several components needed to implement the HAZOP
requirements including: 1) a pneumatic valve, 2) a rupture disk, 3) connecting OPTO to
building alarms.
• Implementation of automated controls in OPTO as specified by the HAZOP review.
• Planning and implementing the refractory curing.
• Developing and implementing a standard operating procedure (SOP).
• Testing and demonstrating the reactor to operate under atmospheric natural gas conditions.
6 SUMMARY AND CONCLUSIONS
75
• Certifying the reactor as a pressure vessel.
• Testing and identifying issues with pressurized natural gas combustion.
• Testing and developing a procedure for pressurized dry-feed coal combustion
As a result of the work completed, the reactor can now be heated using natural gas and
fired with dry pulverized coal. Specific results include:
• The reactor can be run with natural gas and can be monitored remotely from a mobile
application for an indefinite period of time. The reactor was heated overnight on at least
eight occasions.
• The reactor has now been operated multiple times using pulverized dry coal at pressures
up to 16 bar for periods up to 3 hours.
• The reactor program in OPTO has been successfully demonstrated to safely control the
reactor and record data.
• Data have been recorded including wall thermocouples, exhaust thermocouples,
radiometer voltages, and exhaust gas concentrations of O2, CO2, CO, NO, and SO2. These
data have been used to analyze coal feed and burner performance.
The following conclusions can be made for the current reactor performance.
• Reactor preheating is relatively robust. Overnight heating has been repeatedly
demonstrated and is an effective means of preheating the reactor for coal combustion.
• After altering the design of the coal feed system to include a mesh screen for fluidization
and a vibrator to avoid bridging, the coal feed system has demonstrated the ability to feed
coal continuously at reactor pressures of 16 bar and coal feed rates of 122 kW for periods
of 40 minutes. During these periods the coal feeds in a series of high frequency pulses of
unknown mean and fluctuating components.
76
• While the coal feed rate is clearly impacted by the net CO2 flow rate, tests to date have not
been able to demonstrate a correlation between the two suggesting there are variables
influencing the coal flow that are not yet understood.
• When operating under overall fuel lean conditions, the coal flame remains above the
second level reactor window. During periods of fuel rich mixtures, the flame can extend
all the way to the bottom or level 5 window.
• To date, the calculated energy losses in the reactor are not consistent using two different
methods of calculation. Additional work will be needed to reconcile the differences.
• Several reactor systems could be improved to increase reliability. The purge system helps
to keep windows clean but is often overwhelmed by periods of high coal flow rate causing
deposition. The system for removing spray water from the heat exchanger is prone to
plugging. The quench spray can be entrained into the cyclone at high air flow rates. [3]
77
[1] "SHORT-TERM ENERGY OUTLOOK," 6 April 2021. [Online]. Available: https://www.eia.gov/outlooks/steo/report/electricity.php. [Accessed 22 April 2021].
[2] "Data and statistics," [Online]. Available: https://www.iea.org/data-and-statistics?country=WORLD&fuel=Energy%20supply&indicator=ElecGenByFuel. [Accessed 22 April 2021].
[3] "How much carbon dioxide is produced when different fuels are burned?," [Online]. Available: https://www.eia.gov/tools/faqs/faq.php?id=73&t=11.. [Accessed 22 April 2021].
[4] "Coal explained," 9 October 2020. [Online]. Available: https://www.eia.gov/energyexplained/coal/how-much-coal-is-left.php. [Accessed 22 April 2021].
[5] R. Soundaranajan and T. Gundersen, "Coal based power plants using oxy-combustion for CO2 capture: pressurized coal combustion to reduce capture penalty," Applied Thermal Engineering, vol. 61, no. 1, pp. 115-122, 2013.
[6] L. Chen, S. Z. Yong and A. F. Ghoniem, "Oxy-fuel combustion of pulverized coal: Characterization, fundamentals, stabilization and CFD modeling," Progress in Energy and Combustion Science, vol. 38, no. 2, pp. 156-214, 2012.
[7] J. Tuia, "Correlation Pressure, Transport gas Velocities, and Solids Mass Flowrates in High Pressure Fluidized Bed Coal Feed Systems," Master's thesis, Brigham Young University, 2019.
[8] W. Carpenter, "Burner Design for a Pressurized Oxy-Coal Reactor," Master's thesis, Brigham Young University, 2019.
[9] A. B. Skousen, "Design, Fabrication and Testing of a Pressurized Oxy-Coal Reactor Exhaust System," Master's thesis, Brigham Young University, 2019.
[10] J. Harwell, "BYU PRESSURIZED OXY-COMBUSTION REACTOR AND SUB-SYSTEMS," Hazard and Operability Study (HAZOP) Report, 2019.
[11] N. A. Burchfield, "Narrow Angle Radiometer for Oxy-Coal Combustion," Master's thesis, Brigham Young University, 2020.
[12] D. Toporov, P. Bocian, P. Heil, A. Kellermann, H. Stadler, S. Tschunko, M. Forster and R. Kneer, "Detailed investigation of a pulverized fuel swirl flame in CO2/O2 atmopshere," Combustion and Flame, vol. 155, no. 4, pp. 605-618, 2008.
[13] "Pre-Combustion Carbon Capture Research," U.S. Department of Energy, [Online]. Available: https://www.energy.gov/fe/science-innovation/carbon-capture-and-storage-research/carbon-capture-rd/pre-combustion-carbon. [Accessed 14 July 2021].
REFERENCES
78
[14] H. Hagi, M. Nemer, Y. L. Moullec and C. Bouallou, "Towards second generation pulverized coal power plants: Energy penalty reduction potential of pressurized oxy-combustion systems," Energy Procedia, vol. 63 Elsevier Ltd, pp. 431-439, 2014.
[15] "Carbon Capture & Sequestration Technologies," MIT, 30 September 2016. [Online]. Available: https://sequestration.mit.edu/tools/projects/index.html. [Accessed 13 July 2021].
[16] J. Thornock, D. Tobar, D. R. Tree, Y. Xue and R. Tsiava, "Radiative intensity, no emissions, and burnout for oxygen enriched biomass combustion.," Proceedings of the Combustion Institute, vol. 35, pp. 2777-2784, 2015.
[17] L. G. Becker, H. Kosaka, B. Benjamin, S. Doost, R. Knappstein, M. Habermehl, R. Kneer, J. Janicka and A. Dreizler, "Experimental investigation of flame stabilization inside the quarl of an oxyfuel swirl burner," Fuel, vol. 201, pp. 124-135, 2017.
[18] D. Zabrodiec, J. Hees, A. Massmeyer, F. V. Lehn, M. Habermehl, O. Hatzfeld and R. Kneer, "Experimental Investigation of pulverized coal flames in CO2/O2 - N2/O2 - atmospheres: Comparision of solid particle radiative characteristics," Fuel, vol. 201, pp. 136-147, 2017.
[19] S. Li, H. Li, W. Li, M. Xu, E. Eddings, Q. Ren and Q. Lu, "Coal combustion emission and ash formation characteristics at high oxygen concentration in a 1 MWth pilot-scale oxy-fuel circulating fluidized bed," Applied Energy, vol. 197, pp. 203-211, 2017.
[20] C. K. Stimpson, A. Fry, T. Blanc and D. R. Tree, "Line of sight soot volume fraction measurements in air- and oxy-coal flames," Proceedings of the Combustion Institute, vol. 34, no. 2, pp. 2885-2893, 2013.
[21] A. Fry, B. Adams, A. Paschedag, P. Kazalski, C. Carney, D. Oryshchyn, R. Woodside, S. Gerdemann and T. Ochs, "Principles for retorfitting coal burners for oxy-combustion," International Journal of Greenhouse Gas Control, vol. 5, 2011.
[22] S. S. Daood, W. Nimmo, P. Edge and B. M. Gibbs, "Deep-staged, oxygen enriched combustion of coal," Fuel, vol. 101, pp. 187-196, 2012.
[23] Z. Dobó, M. Backman and K. J. Whitty, "Experimental study and demonstration of pilot-scale oxy-coal combustion at elevated temperatures and pressures," Applied Energy, vol. 252, p. 113450, 2019.
[24] Z. Yang, D. Khatri, P. Verma, T. Li, A. Adeosun, B. M. Kumfer and R. L. Axelbaum, "Experimental study and demonstration of pilot-scale, dry feed, oxy-coal combustion under pressure," Applied Energy, vol. 285, p. 116367, 2021.
[25] G. Benelli, D. Cumbo, M. Gazzino and E. Morgani, "Pressurized oxy-combustion of coal with flue gas recirculation: pilot scale demonstration," in Italy: PowerGen Eur. Conf. Exhib., Milan, 2008.
[26] "Sapphire (Al2O3)," MolTech, [Online]. Available: http://www.mt-berlin.com/frames_cryst/descriptions/sapphire.htm. [Accessed 13 May 2021].
[27] H. Suda and S. Yamada, "Force measrements for the movement of a water drop on a surface with a surface tension gradient," Langmuir, vol. 19, no. 3, pp. 529-531, 2003.
[28] J. John and T. Keith, Gas Dynamics, New Jersey: Pearson Education, Inc., 2006.
79
Table A-1. List of high voltage inputs and output (HV-00-0 through HV-04-3).
OPTO Name Description Location Switch_for_Burner_Permission Connects Honeywell Box’s L3 to L6
with 110 VAC power when switched on. HV-00-3
Air_MFC_Solenoid Solenoid valve before MFC 1 (High Pressure Air)
HV-01-0
LP_NG_Solenoid_MFC_2 Solenoid valve before MFC 2 (Low Pressure Natural Gas)
HV-01-1
Spray_Solenoid Spray Quench Solenoid - Fail Open - When NOT energized water passes through
HV-01-2
Emergency_Drain_BV Blow off ball valve (solenoid) for emergency exhaust draining. This valve is in parallel with the badger and bypass valves
HV-01-3
O2_Solenoid_MFC3_MFC4 Solenoid upstream of both oxygen MFCs for the burner. Located on the north side of the reactor below the mezzanine.
HV-02-0
CO2_Solenoid_MFC5_MFC6 Solenoid upstream of both CO2 MFCs for the burner. Located on the north side of the reactor below the mezzanine.
HV-02-1
Ignitor_Power On/Off power switch to the ignitor HV-02-3 Cond_Top_BV_CLOSE Closes the top condensation ball valve. HV-03-0 Cond_Top_BV_OPEN Opens the top condensation ball valve. HV-03-1 Cond_Bottom_BV_CLOSE Closes the bottom condensation ball
valve. HV-03-2
Cond_Bottom_BV_OPEN Opens the bottom condensation ball valve.
HV-03-3
Annulus_Burner_Air_Solenoid Solenoid valve between MFC 1 (High Pressure Air) and the outer annulus.
HV-04-0
APPENDIX A. OPTO I/O POINTS
80
Table A-2. List of high voltage inputs and outputs (HV-05-0 through HV-08-03).
OPTO Name Description Location Electric_Box_Fan ON/OFF to cooling fan on the bottom of
the OPTO box that removes hot air HV-05-0
Electric_Box_CA_Solenoid Solenoid that allows cold air to enter the OPTO box
HV-05-1
Bypass_Pressure_Valve_OPEN Opens the “two-inch” bypass ball valve that controls if the reactor can pressurize
HV-05-2
Bypass_Pressure_Valve_CLOSE Closes the “two-inch” bypass ball valve that controls if the reactor can pressurize
HV-05-3
Fluidized_Coal_BV_Solenoid_CF Bottom solenoid on CF plate. Delivers air to pneumatic valve that separates CF from burner.
HV-06-0
Hopper_BV_Solenoid_CF Second solenoid from bottom on CF plate. Delivers air to pneumatic ball valve on top of coal hopper that is for loading coal.
HV-06-1
Fluidization_CO2_MFC9_Solenoid Second solenoid from top on CF plate. Solenoid that slows CO2 to MFC 9 (fluidization).
HV-06-2
Dilution_CO2_Solenoid_MFC8_CF Top solenoid on CF plate. Solenoid allows CO2 to MFC 8 (Dilution).
HV-06-3
Lance_Burner_Air_Solenoid Separates air MFC from the eight tertiary tubes of the burner.
HV-07-0
Burner_Control_1_On When ON the HP NG solenoid will open if the UV sensors are excited. Used in Ignitor chart to start the reactor.
HV-07-1
Bypass_Valve_Opened_Indicator Opened indicator for bypass valve that controls if the reactor can pressurize. (S4 Series Electric Actuator)
HV-08-2
Bypass_Valve_Closed_Indicator Closed indicator for bypass valve that controls if the reactor can pressurize. (S4 Series Electric Actuator)
HV-08-3
81
Table A-3. List of high voltage inputs and outputs (HV-09-00 through HV-11-03).
OPTO Name Description Location Cond_Top_BV_Closed_Indicator Closed Indicator for TOP Condensation
Ball Valve HV-09-0
Cond_Top_BV_Opened_Indicator Opened Indicator for TOP Condensation Ball Valve
HV-09-1
Cond_Bottom_BV_Closed_Indicator Closed Indicator for BOTTOM Condensation Ball Valve
HV-09-2
Cond_Bottom_BV_Opened_Indicator Opened Indicator for BOTTOM Condensation Ball Valve
HV-09-3
Honeywell_Box_Indicator Signal from Honeywell box if permissions are ON and UV sensors are excited (HP NG solenoid is OPEN)
HV-10-0
Alarm_System_Indicator Alarm system indicator for building gas alarm. Signal from relay in junction box above ERL 106C door.
HV-10-2
Coal_Feed_Pressure_Switch CF pressure switch. Used to hardware lock the ball valve on top of the coal feeder (used for loading the feeder).
HV-11-0
Coal_Feed_Pressure_Switch_NO CF pressure switch. Same as “Coal_Feed_Pressure_Switch” but NORMALLY OPEN.
HV-11-1
Pressure_Switch Located at the top of the cyclone where flue gas exits. Excited when the ID fan is on and a vacuum is being pulled.
HV-11-3
82
Table A-4. List of low voltage inputs (LVI-00-0 through LVI-06-7).
OPTO Name Description Location HX_Flow_Meter On the coolant supply line before the heat
exchanger. Measures coolant flow rate. LVI-00-0
MFC_1_Air_Flowrate Flow rate signal from air MFC 1 (0-5000 SCFH) (outer annulus & tertiary tubes)
LVI-05-0
MFC_3_Oxygen_Flowrate Flow rate signal from oxygen MFC 3 (0-950 SCFH) (tertiary tubes)
LVI-05-2
MFC_4_Oxygen_Flowrate Flow rate signal from oxygen MFC 4 (0-950 SCFH) (inner annulus)
LVI-05-3
Purge_PT Pressure of purge near the gas cylinder rack on the east wall of ERL 106C
LVI-06-0
Coal_Feed_Pressure2 Second pressure transducer on the feed system (installed on the pressure relief line).
LVI-06-1
HP_Air_Pressure_Transmitter HP air pressure transducer on the north wall of ERL 106C where the piping goes under the mezzanine upstream of the MFC.
LVI-06-2
Coal_Line_Pressure_Transmitter First pressure transducer on the coal feed system (installed on the vent line by the CF solenoid and MFC plate).
LVI-06-3
Damper_Cyclone_Dilution_ Pressure_Transmitter
Pressure differential transducer installed at the dilution damper. Must be negative for flue gas to not escape into the room.
LVI-06-4
Cyclone_Pressure_Transmitter Pressure differential transducer installed at top of cyclone where flue gas exits. Measures vacuum being pulled on the system.
LVI-06-6
Spray_Pressure_Transmitter Quench spray pressure transmitter (range 0-2500 psig) located downstream of pump.
LVI-06-7
83
Table A-5. List of low voltage inputs (LVI-07-0 through LVI-08-7).
OPTO Name Description Location CO2_Horiba Horiba CO2 – 4-20 ma Linear self
powered signal LVI-07-0
NOx_Horiba Horiba NOx – 4-20 ma Linear self powered signal
LVI-07-1
SO2_Horiba Horiba SO2 – 4-20 ma Linear self powered signal
LVI-07-2
O2_Horiba Horiba O2 – 4-20 ma Linear self powered signal
LVI-07-3
CO_Horiba Horiba CO – 4-20 ma Linear self powered signal
LVI-07-4
Rupture_Disk_Pressure_Transmitter Pressure transducer install upstream of the rupture disk on the north side of the reactor at waist level.
LVI-08-0
LP_Air_Pressure_Transmitter Building low pressure air pressure transducer (range 0-200 psig). Located above the OPTO control box.
LVI-08-2
HP_NG_Pressure_Transmitter High pressure natural gas pressure transducer installed on the mezzanine in the north-east corner (range 0-500 psig)
LVI-08-3
HX_to_Valves_Pressure_Transmitter Reactor pressure transducer located between the HX and pressure control valves (range 0-500 psig)
LVI-08-4
Burner_Pressure_Transmitter Pressure transmitter located upstream of the burner on the tertiary tubes line.(For reactor pressure differential measurements)
LVI-08-5
O2_Sensor_UofU O2 sensor from the University of Utah - Testing
LVI-08-7
84
Table A-6. List of low voltage inputs (LVI-09-0 through LVI-11-3).
OPTO Name Description Location Control_Damper Control damper where ducting goes above
the mezzanine. (Normally Open) LVI-09-0
Relief_Damper Relief damper above the burner. Controls vacuum pulled on system. (Normally Closed)
LVI-09-1
Spray_Flow_Meter Quench spray flow meter located upstream of the pump.
LVI-09-2
VFD_Flow_Feedback Variable frequency driver for controlling quench spray pump’s RPMs. Located next to the pump on the east wall of ERL 106C.
LVI-09-3
O2_Voltage Zirconium wide band sensor (LSU 4.9). Voltage based on O2 percentage in flue gas. Mounted after pressure control valves.
LVI-10-0
MFC_8_CO2_Flowrate_coalfeed Signal from MFC 8. Controls CO2 for diluting coal transported to the burner. (0-450 SCFH)
LVI-10-1
MFC_9_CO2_Flowrate_coalfeed Signal from MFC 9. Controls CO2 for fluidizing coal. (0-450 SCFH)
LVI-10-2
Dilution_Damper Dilution damper upstream of cyclone. Controls air mixing with flue gas. (Normally Closed)
LVI-10-3
MFC_5_CO2_Flowrate Flow rate signal from carbon dioxide MFC 5 (0-1050 SCFH) (tertiary tubes)
LVI-11-0
MFC_6_CO2_Flowrate Flow rate signal from carbon dioxide MFC 6 (0-1050 SCFH) (inner annulus)
LVI-11-1
MFC_7_HP_Methane_Flowrate Flow rate signal from high pressure natural gas MFC 7 (0-500 SCFH) (inner tube)
LVI-11-2
Load_Cell Signal from three load cells supporting CF. The signals are first processed in a summing box on the ERL 106C east wall.
LVI-11-3
85
Table A-7. List of low voltage outputs (LVO-00-0 through LVO-09-1).
OPTO Name Description Location MFC_1_Air_Setpoint Air MFC 1 flow rate setpoint (range 0-5000
SCFH) (outer annulus & tertiary tubes) LVO-00-0
MFC_5_CO2_Setpoint CO2 MFC 5 setpoint (range 0-1050 SCFH) (tertiary tubes)
LVO-00-1
MFC_3_Oxygen_Setpoint Oxygen MFC 3 setpoint (range 0-950 SCFH) (tertiary tubes)
LVO-01-0
MFC_4_Oxygen_Setpoint Oxygen MFC 4 setpoint (range 0-950 SCFH) (inner annulus)
LVO-01-1
Relief_Damper_Setpoint Setpoint for Relief Damper above the burner. Controls vacuum pulled on the system (Normally Closed)
LVO-02-0
VFD_Control_Setpoint Setpoint for the quench spray pump’s RPMs. Located next to the pump on the east wall of the ERL 106C.
LVO-02-1
Dilution_Damper_Setpoint Setpoint for Dilution Damper upstream of cyclone. Controls air mixing with flue gas. (Normally Closed)
LVO-03-0
Control_Damper_Setpoint Setpoint for Control Damper where ducting goes above the mezzanine. (Normally Open)
LVO-03-1
Pressure_Control_Valve_Setpoint Pneumatic Badger Valve for fine tuning reactor pressure (4.64 to 7.2 mA is 0% to 100% open)
LVO-04-1
MFC_8_CO2_Setpoint_Coalfeed CF CO2 MFC 8 setpoint (range 0-450 SCFH) (dilution)
LVO-08-0
MFC_6_CO2_Setpoint CO2 MFC 6 setpoint (range 0-450 SCFH) (inner annulus)
LVO-08-1
MFC_9_CO2_Setpoint_Coalfeed CF CO2 MFC 9 setpoint (range 0-450 SCFH) (fluidization)
LVO-09-0
MFC_7_HP_Methane_Setpoint HP NG MFC 7 setpoint (range 0-500 SCFH) (inner tube)
LVO-09-1
86
Table A-8. List of low voltage outputs (LVO-11-0 through LVO-15-0).
OPTO Name Description Location Red_Light On/Off for red alarm light located above
ERL 106C door LVO-11-0
Green_Light On/Off for green alarm light located above ERL 106C door
LVO-11-1
Yellow_Light On/Off for yellow alarm light located above ERL 106C door
LVO-11-2
O2_Sensor_OnOff On/Off for zirconium oxygen sensor LVO-11-3 MFC_Test_Out Temporary for MFC setpoint testing LVO-14-0 MFC_Test_In Temporary for MFC input signal testing LVO-15-0
87
Table A-9. List of thermocouple inputs (TCI-00-0 through TCI-01-3).
OPTO Name Description Location Therm_B_Inner_Top Type U/B Thermocouple Located at the Top
of the reactor; closest to the refractory inner wall (embedded)
TCI-00-0
Therm_B_Outer_Top Type U/B Thermocouple Located at the Top of the reactor; furthest from the refractory inner wall (embedded)
TCI-00-1
Therm_B_Inner_Middle Type U/B Thermocouple Located 3rd from the Bottom of the reactor; closest to the refractory inner wall (embedded)
TCI-00-2
Therm_B_Outer_Middle Type U/B Thermocouple Located 3rd from the Bottom of the reactor; furthest from the refractory inner wall (embedded)
TCI-00-3
Therm_B_Inner_MiddleBottom Type U/B Thermocouple Located 2nd from the Bottom of the reactor; closest to the refractory inner wall (embedded)
TCI-00-4
Therm_B_Outer_MiddleBottom Type U/B Thermocouple Located 2nd from the Bottom of the reactor; furthest from the refractory inner wall (embedded)
TCI-00-5
Therm_B_Inner_Bottom Type U/B Thermocouple Located at the Top of the reactor; closest to the refractory inner wall (embedded)
TCI-00-6
Therm_B_Outer_Bottom Type U/B Thermocouple Located at the Top of the reactor; furthest from the refractory inner wall (embedded)
TCI-00-7
SS_Connector_Therm_Reactor_TC K Thermocouples for Stainless Steel Connector held by a band clamp (On the reactor’s side) (surface)
TCI-01-0
SS_Connector_Therm_HX_TC K Thermocouple for the Stainless Steel Connector held by a band clamp (On the heat exchanger’s side) (surface)
TCI-01-1
HX_Flue_Gas_IN_TC K thermocouple where flue enters the heat exchanger (probe)
TCI-01-2
HX_Flue_Gas_OUT_TC K thermocouple where flue exits the heat exchanger (probe)
TCI-01-3
88
Table A-10. List of thermocouple inputs (TCI-01-4 through TCI-03-7).
OPTO Name Description Location HX_Cold_Water_Out_TC K thermocouple where coolant exits the
Heat Exchanger (probe) TCI-01-4
HX_Cold_Water_In_TC K thermocouple where coolant enters the Heat Exchanger (probe)
TCI-01-5
Therm_B_Inner_MiddleTop Type U/B Thermocouple Located 4th from the Bottom of the reactor; closest to the refractory inner wall (embedded)
TCI-01-6
Therm_B_Outer_MiddleTop Type U/B Thermocouple Located 4th from the Bottom of the reactor; furthest from the refractory inner wall (embedded)
TCI-01-7
Magnet_6 Magnet 6 K thermocouple on reactor’s shell (surface)
TCI-02-0
Magnet_5 Magnet 5 K thermocouple on reactor’s shell (surface)
TCI-02-1
HX_SS310_Weld_TC Located at the weld where flue gas enters the HX. Held to the weld by a band clamp (surface)
TCI-02-3
Magnet_2 Magnet 2 K thermocouple on reactor’s shell (surface)
TCI-02-4
Magnet_1 Magnet 1 K thermocouple on reactor’s shell (surface)
TCI-02-5
Opto_Box_Therm K thermocouple inside the OPTO control box
TCI-02-6
Spray_Flue_Gas_Therm K thermocouple slightly upstream of the quench spray nozzle. The nozzle introduces water downstream towards the HX (probe)
TCI-02-7
Magnet_3 Magnet 3 K thermocouple on reactor’s shell (surface)
TCI-03-2
Blow_Off_Valve_TC K thermocouple located on the unpressurized side of the rupture disk (probe)
TCI-03-4
Magnet_4 Magnet 4 K thermocouple on reactor’s shell (surface)
TCI-03-5
Magnet_8 Magnet 8 K thermocouple on reactor’s shell (surface)
TCI-03-6
Magnet_7 Magnet 7 K thermocouple on reactor’s shell (surface)
TCI-03-7
89
Pretest Checklist
� On the Pad
o Turn on the High Pressure Air Compressor (takes about 30 minutes to pressurize
tank)
o Open the High Pressure Natural Gas shutoff valve
� Remove all unnecessary equipment left on or near the reactor
� Close the OPTO Box
� Put up warning sign near entrance of room ERL 106C
� Turn on AMCREST Camera
o Open “AMCREST IP CONFIG” (desktop application) and note current IP address
o Enter http://(Ip_Address)
Username: admin
Password: video2POC
� Radiometers
o Turn on chiller
o Verify water flow through chiller lines
� Horiba PG-250
o Disconnect stainless steel tubing from flange port
o Turn On Horiba
o After a few minutes hit the exit button
o Verify Calibrations
For air (21% O2)
o Make sure water is above inner tube
o Connect stainless steel tubing to flange port
APPENDIX B. STANDARD OPERATING PROCEDURE
90
o Open Needle valve to reactor
� Zirconium Oxygen Sensor
o Verify Vacuum Pump Exit Tube is connected to the Dilution Damper
o Verify Vacuum Pump is Plugged In
o Open Dilution Damper to 5%
o Turn on O2 Sensor Power
� Spray Quench
o Close the Spray Quench Solenoid valve
o Open the Quench Water supply ball valve
o Turn on the VFD
If needed reset the breaker in the southern breaker box in 106C (breaker 1)
Set VFD Code to 00-05 and range to 02 (reference L510 pg. 4-9)
� Condensate Pump
o Connect the plug to the wall outlet
o Position the pan and bucket under the condensate valves
o Connect the plastic tube to the floor drain
� Heat Exchanger Coolant
o Open the inlet and outlet coolant valves (PCHWR & PCHWS)
o Verify that the cooling water bypass is closed
o Check HX cooling water flow, “Coolant”, on OPTO 22
� Ventilation
o Turn on the room ventilation (switch outside ERL 106c door)
o Turn on ID fan
Set to hand
o Turn on the Bypass Valve Interlock Chart (OPTO 22)
Open the Bypass Valve
Verify Indicator is Open
o Turn on the Badger Valve Chart (OPTO 22)
Set Pressure Control Valve to 100%
o Dampers
Verify relief dampers is closed
91
Verify that the control damper is open
o Verify that the reactor pressure is negative (check pressure switch is on)
� Purge System
o Connect K cylinder to the purge line (Argon, N2, CO2)
o Drain rupture disk ball valve
o Verify Desired purge line ball valves are open
o Open the purge cylinder and set purge regulator to pressure to 150-200 psig.
purge flow
o Open purge ball valve
o Verify purge pressure on OPTO
� Turn on Charts
o System Interlock Chart
Click AUTO button (Cycles the ball valves)
o Heat Flux and Inner Temperature Chart (If B thermocouples are displaying
signal)
o Control Box Cooling Chart
o Quench Spray VFD Chart (OPTO 22)
� Pressurized Air
o Open the High Pressure Air Ball Valve
o Open the High Pressure Air Ball Valve Next to the MFC
o Verify both High Pressure Air Regulator
Verify that the High Pressure Air is above 80 psig
o Verify the Low Pressure Air is above 50 psig
� Ignitor
o Remove UV Sensor flange from northeast “half-inch” flange (on mezzanine)
Unscrew the four bolts
Unscrew the Swagelok connection to the purge line
Remove UV Sensor flange and gasket
o Insert the Ignitor until the black stopper rests on “half-inch” flange
Turn the ignitor so the line on the black stopper is aligned with the black
checkmark on the flange
92
� High Pressure Natural Gas
o Physically verify that the Fluidized coal ball valve is closed
o Open High Pressure Natural Gas Shutoff Valve and Ball Valve in ERL 106C
o Verify the High Pressure Natural Gas Regulator
Do not exceed 500 psig. (if above 500 psig. unscrew regulator slightly.
The gauge won’t reflect this change until there is Natural Gas flow)
o Open HP NG Ball Valve between burner and MFC
Testing Procedure
� Turn on High Pressure Air Solenoid
o Turn on Lance Solenoid
o Turn on Outer Annulus Solenoid
� Open Excel Sheet (CH4_FA_ratio_V2.xlsx)
o “Air Operation” Sheet
� Set air flow to 61 kg/hr (~40kW)
o Use “O2 3% (Vol)” column for selecting air firing rates
o Verify air flow stabilizes at startup set point
� Check Honeywell Burner Control Module to reset permissions
� Start the Ignitor Chart (OPTO 22)
� Setting Natural Gas Flow Rate
o Verify Natural Gas Solenoid is Open
o Verify Natural Gas Flow Rate is 0 kg/hr
o Set the Natural Gas Flow Rate to 3 kg/hr (~40kW)
o Verify Natural Gas Flow Stabilizes
� Check O2 Sensor Voltage reading
o Make sure Voltage fluctuates at 0.3 Volts
� Turn on Emergency Trigger Logic Chart (OPTO 22)
o Verify reactor doesn’t trigger shutdown
� Turn off Ignitor (OPTO 22)
o Verify the UV detectors can see the flame and that the Honeywell box
permissions is not tripped
� Turn on Quench Spray solenoid (OPTO 22)
93
� Verify reactor pressure is negative
� Remove Ignitor and attach the UV Sensor flange
o Torque flange to 35 ft-lbs.
o Attach UV detector
o Attach the Purge Line
o Open the Purge Line Ball Valve
� Turn on Data Collection (OPTO 22)
o Verify files open are saving data
� Turn on historical data log for radiometers
Running Test
� Operator has an air fuel ratio calculator (CH4_FA_ratio_V2.xlsx) and maintains
stoichiometric ratios
� If heating up to introduce coal
o Refractory is heated to be at least 1900 oF throughout (preferably 2400 oF the
minimum slagging temperature)
o Pressurized Coal Checklist on following page is followed before introducing coal
Shutdown Procedure
� Start Normal Shutdown Chart (OPTO 22)
� Verify that the Normal Shutdown Chart starts (OPTO 22)
� Allow air to purge until O2 Sensor voltage returns to above 5 Volts
� Set air flow to zero
� Close High Pressure Air Solenoid
� Close High Pressure Natural Gas Shutoff Valves in ERL 106C (three valves in total)
Post-Shutdown Procedure
� Drain rupture disk ball valve
� Go through the Pretest Checklist and reverse everything done
� Turn off the red emergency light
� Leave fan on Stainless Steel Connector turned on
� Leave Heat Exchanger Coolant running
94
Pressurized Coal Checklist
1. On the Pad
a. Turn on CO2 Heater
i. Slowly open CO2 ball valve mounted on wall
b. Slowly open O2 ball valve mounted on the wall
2. Spray down Coal Hopper Ball Valve with compressed air
3. Turn on Coal Mass Flow Chart
4. Tare Coal feeder
5. Fill barrel half full (30.5 cm (12 in) in barrel = a full hopper)
6. Fill the hopper with coal
7. Spray down Coal Hopper Ball Valve with compressed air
8. Close the hopper ball valve
9. HPNG set point set to 0 kg/hr
10. Close 3 HPNG Ball valves on mezzanine
11. Turn on fan to cool Stainless Steel Connector (fan placed directly underneath Stainless
Steel connector)
12. Drain rupture disk ball valve
13. Increase purge line pressure (500 psig)
14. Continually adjust Horiba flow rate (this is done by adjusting the needle valve)
15. High Pressure Carbon Dioxide in ERL 106C on the mezzanine
a. Unscrew the regulator out completely to stop flow
b. Slowly open ball valve
c. Slowly screw the regulator until the back pressure gauge reads 250 psig.
16. High Pressure Oxygen in ERL 106c on the mezzanine
a. Unscrew regulator out completely to stop flow
b. Verify Needle valve downstream of the pressure regulator is open
c. Slowly open the ball valve
d. Slowly screw regulator in until the back pressure gauge reads 250 psig.
17. Turn off coal feeder vent (rotameter)
18. Turn on CO2 to the coal feeder (start pressurizing feeder)
19. Close Bypass Valve
95
20. Open Badger valve to 60% (continually adjust)
21. Pressurize reactor with air to 100 psig
22. Close 3 high pressure air line solenoids (outer annulus, lance, MFC)
23. Close air ball valve by MFC
24. Turn on CO2 to reactor
25. Increase CO2 Pressure Regulator as needed
26. Let pressure drop (dp = hopper pressure – reactor pressure) be between 3-8 psi
27. Turn on O2 to desired amount
28. Increase O2 Pressure Regulator as needed
29. With pressure drop between 3-8 psi open burner coal BV
(100 kW) Pressurized Coal - Mass Flow Controller (MFC) Flow rates
Flow rates were selected from Cody Carpenter’s thesis. In his thesis three cases were
modeled. The third case results in the desired temperature profile in the reactor. The third case
flow rates are shown below.
Table B-1. Burner MFC nominal flow rates from case 3 for a firing rate of 100 kW.
Location Fluid Flow Rate (kg/hr) Outer Annulus/Tertiary Tubes Air 0.0 N/A Low Pressure Natural Gas (LP NG) 0.0 Tertiary Tubes O2 26.9 Inner Annulus O2 3.2 Tertiary Tubes CO2 26.9 Inner Annulus CO2 12.6 Primary Tube High Pressure Natural Gas (HP NG) 0.0
Table B-2. Coal feed MFC nominal flow rates from case 3 for a firing rate of 100 kW.
Location Fluid Flow Rate (kg/hr) Primary Tube Dilution CO2 9.7 Fluidization CO2 4.0 Primary Tube Coal 12.3 Vent line [rotameter] CO2 1.5 [50th notch]
96
List of hardware on the POC Reactor and POC Coal Feeder that require regular
maintenance. The following bullet point list contains generic protocol that should be followed
when working on the POC system.
• Snoop all fittings that have been disconnected or tampered with
• Clean unspecified reactor filters every 3 months
Table C-1. Coal Feed hardware maintenance.
Location Hardware Frequency Action/Comments Vent line. Mounted near the top of the hopper next to the mezzanine.
Black Filter Every Month Clean out the filter.
Pressure Relief Line. Laying on the floor by the base of the hopper.
Black Filter Every Month Clean out the filter.
Between Dilution CO2 MFC and hopper
Particle Filter
Every 3 Months Clean out the filter
Between Fluidization CO2 and fluidization mechanism
Particle Filter
Every 3 Months Clean out the filter
On top of the hopper. Where coal is loaded into the hopper.
Ball Valve After filling hopper with coal
Spray off the ball valve with compressed air before closing it. Valve may not seal If this step is not taken.
Coal line Tubing Every 6 Months Inspect coal line for corrosion
APPENDIX C. REGULAR MAINTENANCE LIST
97
Table C-2. Reactor hardware maintenance.
Location Hardware Frequency Action/Comments Purge Line Nitrogen K-
Cylinders Before Every Run
Have BYU Central Stores replace the cylinders. BYU Central Stores: (801)422-6444 Brent’s Cell Phone: (801)360-5430
Spray Quench Line
Pump Every 1000 hours of operation
Replace the oil. Oil: Cat Pumps - Crankcase oil ISO 68 Special Formula Premium Grade
“2-inch” Viewing Ports
Sapphire Windows
Whenever Dirty Clean deposits on the window makes optical measurements impossible
Purge Line Orifices Whenever Clogged
Clean When orifices are clogged. Purge K-cylinders will drain slower when orifices are clogged.
Spray Quench Line
Spray Quench Nozzle
After Every Run
Deposits can cause the water to not atomize or spray at all
Directly after the Heat Exchanger
Pressure Control Valve
Every 3 Months Deposits will slowly build up and must be cleaned. This valve I also known as the badger valve.
Piping above the mezzanine between the Emergency Drain and Rupture Disk
Black pipe under the tee
Every 6 Months Clean this section of pipe. Also known as the “little leg”.
Combustion Chamber
Refractory Every 6 Months Schedule inspection
Reactor Shell P&ID (E-1) Every 6 Months Inspect Combustion Chamber
Slag pit Every 6 Months Clean out slag that builds up. Wear steel toed shoes. Use transmission jack system.
98
D.1 Sample Identification: #2
Date: 3/26/2021
Table D-1. Mineral Ash Analysis
Parameter Percent Weight Ignited Basis Silica, SiO2 64.41
Alumina, Al203 15.16 Titania, TiO2 0.74
Lime, CaO 4.90
Ferric Oxide, Fe2O3 4.59 Potassium Oxide, K2O 0.75
Magnesium Oxide, MGO 0.99 Sodium Oxide, Na2O 2.27
Sulfur Trioxide, SO3 5.17
Phosphorus Pentoxide, P2O5 0.71 Barium Oxide, BaO 0.06
Manganese Dioxide, MnO2 0.01 Strontium Oxide, SrO 0.13
Undetermined 0.11
Alkaline Content, Total on Ash 0.25 Silica Value 86.01
Base to Acid Ratio 0.17 Estimated T250 Temperature 2500
APPENDIX D. MINERAL ASH, ULTIMATE, AND PROXIMATE ANALYSES
99
Table D-2. Ultimate Analysis
% As Received Basis Dry basis Moisture 3.03 N/A
Carbon 68.56 70.7 Hydrogen 4.96 5.12 Nitrogen 1.47 1.52
Ash 8.62 8.89 Sulfur 0.56 0.58
Oxygen 12.79 13.19 100.00 100.00
100
Table D-3. Proximate Analysis
As Received Dry Basis % Moisture 3.03 N/A
% Ash 8.62 8.89 % Volatile Matter 69.59 71.86
% Fixed Carbon 18.76 19.35 100.00 100.00
% Sulfur 0.56 0.58 Btu/Lb. 12395 12782
Moisture Ash Free Btu/Lb. 14030
101
D.2 Sample Identification: #3
Date: 3/26/2021
Table D-4. Mineral Ash Analysis
Parameter Percent Weight Ignited Basis Silica, SiO2 61.43
Alumina, Al203 16.22 Titania, TiO2 0.81
Lime, CaO 5.32
Ferric Oxide, Fe2O3 4.76 Potassium Oxide, K2O 0.78
Magnesium Oxide, MGO 1.05 Sodium Oxide, Na2O 2.55
Sulfur Trioxide, SO3 5.33
Phosphorus Pentoxide, P2O5 0.81 Barium Oxide, BaO 0.07
Manganese Dioxide, MnO2 0.01 Strontium Oxide, SrO 0.14
Undetermined 0.72
Alkaline Content, Total on Ash 0.24 Silica Value 84.66
Base to Acid Ratio 0.18 Estimated T250 Temperature 2500
102
Table D-5. Ultimate Analysis
% As Received Basis Dry basis Moisture 2.91 N/A
Carbon 69.09 71.16 Hydrogen 5.00 5.15 Nitrogen 1.47 1.51
Ash 7.72 7.95 Sulfur 0.55 0.57
Oxygen 13.27 13.66 100.00 100.00
103
Table D-6. Proximate Analysis
As Received Dry Basis % Moisture 2.91 N/A
% Ash 7.72 7.95 % Volatile Matter 41.98 43.24
% Fixed Carbon 47.39 48.81 100.00 100.00
% Sulfur 0.55 0.57 Btu/Lb. 12592 12969
Moisture Ash Free Btu/Lb. 14090
104
D.3 Sample Identification: #4
Date: 3/26/2021
Table D-7. Mineral Ash Analysis
Parameter Percent Weight Ignited Basis Silica, SiO2 61.28
Alumina, Al203 15.46 Titania, TiO2 0.79
Lime, CaO 7.82
Ferric Oxide, Fe2O3 4.47 Potassium Oxide, K2O 0.63
Magnesium Oxide, MGO 0.96 Sodium Oxide, Na2O 2.35
Sulfur Trioxide, SO3 5.06
Phosphorus Pentoxide, P2O5 0.71 Barium Oxide, BaO 0.07
Manganese Dioxide, MnO2 0.02 Strontium Oxide, SrO 0.13
Undetermined 0.25
Alkaline Content, Total on Ash 0.24 Silica Value 82.22
Base to Acid Ratio 0.21 Estimated T250 Temperature 2500
105
Table D-8. Ultimate Analysis
% As Received Basis Dry basis Moisture 3.09 N/A
Carbon 68.26 70.44 Hydrogen 4.97 5.13 Nitrogen 1.47 1.52
Ash 8.43 8.70 Sulfur 0.55 0.57
Oxygen 13.22 13.64 100.00 100.00
106
Table D-9. Proximate Analysis
As Received Dry Basis % Moisture 3.09 N/A
% Ash 8.43 8.70 % Volatile Matter 41.59 42.92
% Fixed Carbon 46.89 48.38 100.00 100.00
% Sulfur 0.55 0.57 Btu/Lb. 12430 12826
Moisture Ash Free Btu/Lb. 14049
107
D.4 Sample Identification: #6
Date: 3/26/2021
Table D-10. Mineral Ash Analysis
Parameter Percent Weight Ignited Basis Silica, SiO2 65.34
Alumina, Al203 15.32 Titania, TiO2 0.76
Lime, CaO 5.24
Ferric Oxide, Fe2O3 3.57 Potassium Oxide, K2O 0.74
Magnesium Oxide, MGO 1.27 Sodium Oxide, Na2O 1.35
Sulfur Trioxide, SO3 5.11
Phosphorus Pentoxide, P2O5 0.63 Barium Oxide, BaO 0.08
Manganese Dioxide, MnO2 0.01 Strontium Oxide, SrO 0.12
Undetermined 0.46
Alkaline Content, Total on Ash 0.20 Silica Value 86.63
Base to Acid Ratio 0.15 Estimated T250 Temperature 2500
108
Table D-11. Ultimate Analysis
% As Received Basis Dry basis Moisture 2.60 N/A
Carbon 67.47 69.27 Hydrogen 4.89 5.02 Nitrogen 1.39 1.43
Ash 10.50 10.78 Sulfur 0.48 0.49
Oxygen 12.67 13.01 100.00 100.00
109
Table D-12. Proximate Analysis
As Received Dry Basis % Moisture 2.60 N/A
% Ash 10.50 10.78 % Volatile Matter 40.27 41.35
% Fixed Carbon 46.63 47.87 100.00 100.00
% Sulfur 0.48 0.49 Btu/Lb. 12212 12538
Moisture Ash Free Btu/Lb. 14053
110
E.1 Coal Mass Flow
The “Coal_Mass_Flow” chart, shown in Figure E-1, monitors and controls the coal feeder
subsystem. When starting the chart, a table is populated with the voltage signals from the load
cells. The “Coal_Time_Differential_Setpoint” is a numeric variable that controls how long the
table takes to populate and thereby how many data points are stored in the table. The default for
the “Coal_Time_Differential_Setpoint” is 30 seconds. Each load cell voltage is averaged over one
hundred data points before being added to the mass table. The one hundred data points are averaged
in less than a second. Simultaneously, a time chart is populated with the time stamp that the voltage
table is populated. Once these tables are populated the slope of a linear regression line for the mass
table with respect to the time table is calculated with Equation (E-1).
𝑚𝑚 =
𝑁𝑁(∑ xy) − (∑ x)(∑𝑦𝑦)𝑁𝑁(∑𝑥𝑥2) − (∑𝑥𝑥)2 (E-1)
This slope is the calculated coal mass flow rate. “Coal_Time_Differential_Setpoint” is
initialized to thirty seconds. Thirty seconds was found experimentally to maximizing data points
for accurate calculations, while minimizing the time needed for a data point to cycle through the
entire table.
This chart also contains a software lock for the solenoid,
“Fluidized_Coal_BV_Solenoid_CF”, that controls flow from the coal feeder to the reactor.
APPENDIX E. OPTO PAC CONTROL (CHARTS)
111
Currently, the solenoid is programmed to open only if the pressure of the coal feed is between 21
kPa (3 psi) and 55.2 kPa (8 psi) more than the reactor. This is to prevent backflow of hot reactor
gases from flowing from the reactor back into the coal feeder.
Figure E-1. Coal feed chart flow chart.
E.2 Pressure Control Valve
The “Badger_Valve” chart, shown in Figure E-2, controls the position of the Pressure
Control Valve which controls the reactor pressure. This charge also determines if the Pressure
Control Valve opening is manually or PID controlled. If the chart is off the valve position must be
changed manually. When the chart is on, PID communication can be turned on along with the PID
set to auto to enable the PID to automatically adjust the badger valve. Then, the desired reactor
pressure can be selected within the PAC display and the PID will automatically adjust the Pressure
Control Valve.
112
Figure E-2. Badger valve flow chart.
E.3 Bypass Valve Interlock
The “Bypass_Valve_Interlock” chart, shown in Figure E-3, monitors the bypass valve and
controls the alarm lights. The “two-inch” ball valve has a software lock that does not allow the
valve to open if the reactor pressure is above 69 kPa (10 psig). If the reactor pressure does exceed
69 kPa (10 psig) the yellow alarm light turns on above the ERL 106C door. This lets individuals
in the lab know that the reactor is pressurized and that they should not enter unless directed to. At
pressures below 10 psig in the reactor the chart turns on the green alarm light.
113
Figure E-3. Bypass valve chart.
E.4 System Interlock
The “System_Interlock” chart, shown in Figure E-4, controls the two condensation ball
valves. These valves are located below the heat exchanger and are designed to remove
condensation from the system. Configured in a hatch system, only one ball valve opens at a time
to ensure that the reactor is never open to the room. The chart allows for auto and manual controls.
During automated operation the ball valves cycle with a delay that can be set in the PAC Display.
The delay is initialized to be thirty seconds. The condensation ball valves have closed and open
indicators, the closed indicators of the top ball valve must be on for the bottom ball valve to open.
When set to automatic or manual the chart does not allow both ball valves to be open. To open
114
both ball valves the chart must be turned off and the I/O points have to be opened in the PAC
Controller.
Figure E-4. Condensate ball valve chart.
E.5 Control Box Cooling
The “Control_Box_Cooling” chart, shown in Figure E-5, monitors the temperature of the
OPTO control box. A thermocouple is mounted inside the control box to monitor the temperature
of the electrical components. If the temperature reads above 30oC (85oF), a solenoid will open
allowing cool air to flow into the box. The air passes through a cyclone separator producing a cold
and hot stream. The cool stream enters the box through a tube. This tube has small holes drilled
periodically to allow distributed air to enter the box without producing a high velocity jet. At the
same time the solenoid opens, a fan mounted on the bottom of the box turns on removing air from
115
the box. A deadband was created to prevent the solenoid from opening and closing in rapid
succession as the temperature oscillated around 30oC (85oF). The deadband is a temperature range
that does not change the solenoid state. Because of the deadband, once the solenoid opens the
thermocouple temperature must drop below 26.7oC (80oF) for the solenoid to close.
Figure E-5. OPTO control box chart.
E.6 Emergency Trigger Logic
The “Emergency_Trigger_Logic” chart monitors the reactors input sensors and decides if
the system should shut down. Due to the size of the flowchart, it is shown in two images, Figure
E-6 and Figure E-7, to allow the text to be readable. Temperatures, oxygen concentrations,
pressures, quench spray pressure, GROOV, and the flame detector are monitored.
116
Upon starting the chart, the threshold values for various triggers are first processed by the
code within the grey boxes. These threshold values are shown in Table E-1. Each box contains
code that performs calculations or compares input signals to thresholds. Any needed calculations
are performed by the green box’s code. Thermocouples are checked to be below predetermined
thresholds by the purple boxes. The zirconium oxygen sensor is check to be below 21 percent
oxygen by the magenta box. The dark green box checks if the GROOV shutdown has been
activated. Light green boxes then check the reactor pressures and the air pressures needed to
control valves. The yellow box checks the quench spray pressure to make sure the pressure doesn’t
exceed the systems max pressure of 13.8 MPa (2000 psig).
117
Table E-1. “Emergency_Trigger_Logic” chart thresholds to decide if reactor shutdowns.
I/O Point Threshold Logic Spray_Pressure_Upper_Limit 2000 Quench spray pressure must be
below 2000 psig Reactor_Pressure_Upper_Limit 320 Reactor pressure must be below
320 psig Rupture_Disk_Temp_Upper_Limit 450 Rupture disk temperature must be
below 450 oF Reactor_Shell_Surface_Temp_Upper_Limit 400 Reactor shell temperature must be
below 400 oF Refractory_Temp_Limit 3400 Refractory temperature must be
below 3400oF SS_Connector_Temp_Upper_Limit 938 Stainless steel connector must be
below 938oF Reactor_dp_Upper_Limit N/A The maximum allowable pressure
differential for the pressure transducers still needs to be set
HX_Outlet_Temp_Upper_Limit 400 Heat exchanger surface temperature must be below 400 oF
Coolant_Temp_Upper_Limit 180 Heat exchanger coolant must be below 180oF
Opto_Box_Temp_Upper_Limit 90 Control box temperature must be below 90oF
O2_Upper_Limit 5 Zirconium oxygen sensor signal must be below 5 V
LP_Air_Pressure_Lower_Limit 50 Low pressure air supply pressure must be above 50 psig
HP_Air_Pressure_Lower_Limit 75 High pressure air supply pressure must be above 75 psig
A numeric table, “Emergency_Logic_Unpressurized_Vector” stores a value for each box
in the flowchart. If any box is outside of the predetermined operating conditions an index of the
table is incremented and a message is displayed to the user explaining which operating condition
is not being met. After a one second delay the chart cycles through the same boxes checking
threshold against reactor inputs. If a box is tripped three times in a row the
“Emergency_Shutdown_Procedure” chart is started.
118
Figure E-6. Top section of the Emergency Trigger Logic Chart.
119
Figure E-7. Bottom section of the Emergency Trigger Logic Chart.
120
E.7 Emergency Shutdown Procedure
The “Emergency_Shutdown_Procedure” chart, shown in Figure E-8, contains the
procedure that is executed when a shutdown is triggered (see section 0) . A shutdown triggers the
red alarm light above the ERL 106C door is to go on when this chart is activated. The
“Emergency_Trigger_Logic” chart is also turned off to stop the shutdown procedure from
occurring twice. After these initial steps, one of two paths is taken, either the pressurized or
atmospheric procedure is executed depending on the pressure of the reactor. Both procedures turn
off the burner mass flow controllers and Honeywell box’s permissions. The atmospheric procedure
turns off the air and natural gas mass flow controllers. The pressurized procedure shuts off every
mass flow controller except CO2 introduced through the burner’s tertiary and inner annulus tubes.
CO2 is introduced to purge and cool the reactor, but CO2 flow rates are decreased if the reactor is
increasing pressure.
Figure E-8. Shutdown flow chart.
121
E.8 Heat Loss and Inside Temperature
The “Heat_Flux_and_Inside_temperature” chart, shown in Figure E-9, makes calculations
based on the signals from the B thermocouples embedded in the reactor’s refractory approximately
11.43 cm (4.5 in) and 16.51 cm (6.5 in) from the center of the reactor. These calculations assume
the B thermocouples are located 11.43 cm (4.5 in) and 16.51 cm (6.5 in) from the center of the
reactor. The heat loss and inner wall temperature of the refractory are calculated in the cylindrical
coordinate system. The refractory inner wall has a radius of 10.16 cm (4 in). The thermal
conductivity for the refractory, UltraGreen (SR), was found from the supplier’s website. The
conductivity is a function of temperature. The mean temperature of two B thermocouples are found
and a curve fit used to calculate the conductivity is shown in Equation (E-2).
k �
𝐵𝐵𝑅𝑅𝐵𝐵ℎ𝑟𝑟 ∗ 𝑖𝑖𝑖𝑖 ∗ 𝐹𝐹
� =23.980055473 ∗ 𝑅𝑅[𝐹𝐹]𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚−0.0415003477
144 (E-2)
With the thermal conductivity found the heat flux and inside surface temperature can be
found with the following equations. The heat flux is converted to the heat loss of a 30.5 cm (1 ft)
section of refractory in Equation ((E-3) by assuming the axial temperature profile is constant for
the 15.2 cm (6 in) section of refractory above and below a pair of thermocouples. This conversion
from heat flux to heat loss is done to make the calculated results easier to understand.
𝑄𝑄 �
𝐵𝐵𝑅𝑅𝐵𝐵ℎ𝑟𝑟
� =(𝑅𝑅𝑖𝑖𝑚𝑚 − 𝑅𝑅𝑜𝑜𝑜𝑜𝑜𝑜)𝑘𝑘(24π)
𝑙𝑙𝑖𝑖( 𝑟𝑟𝑜𝑜𝑜𝑜𝑜𝑜𝑟𝑟𝑖𝑖𝑚𝑚)
(E-3)
122
𝑅𝑅𝑤𝑤𝑚𝑚𝑤𝑤𝑤𝑤[𝐹𝐹] =
𝑄𝑄 𝑙𝑙𝑖𝑖 � 𝑟𝑟𝑖𝑖𝑚𝑚𝑟𝑟𝑤𝑤𝑚𝑚𝑤𝑤𝑤𝑤�
𝑘𝑘(24π) + 𝑅𝑅𝑖𝑖𝑚𝑚 (E-4)
Every second a new inner refractory heat loss and inner wall temperature is calculated.
These values are stored in numeric tables and are shown on the display.
Three of the inner B thermocouples do not work. For this reason, heat loss calculations
cannot be calculated for three of the B thermocouple pairs. If an inner B thermocouple is displaying
an error message, the heat loss and inner wall temperatures will similarly display an error message
Figure E-9. Heat and temperature calculation chart.
123
E.9 Ignitor
The “Ignitor” chart, shown in Figure E-10, automates the procedure needed to open the
high-pressure natural gas solenoid. There are two UV detectors to measure emission from the
flame. This solenoid is hardware locked by the Honeywell system if these UV detectors are not
excited by emission from the ignitor of from a flame. For the solenoid to open, the Honeywell
permissions must be turned on, the ignitor activated, and after at least a four-second delay the
UV sensors must be excited by UV irradiation. The solenoid will open to natural gas flow if
either of the two detectors is excited. The Honeywell box must be reset and the same procedure
followed for the solenoid to reopen.
Figure E-10. Ignitor flow chart.
124
E.10 Quench Spray VFD
The “Quench_Spray_VFD” chart, shown in Figure E-11, controls the spray quench pump’s
rotational speed which controls the spray flow rate. The stainless steel connector temperature is
checked every twenty seconds; if the temperature increases or decreases by a degree, the pump’s
RPMs increase or decrease. Water is continually introduced during operation to keep the spray
nozzle from being destroyed. This chart has controlled the pump during many overnight heat ups
and shown that it can control the stainless steel connector’s temperature to be below the threshold
of 503oC (938 oF).
Figure E-11. Quench spray flow chart.
125
During testing, two subsystems of the reactor have produced ongoing issues that have
been partially resolved but continue to hamper successful testing. These subsystems are the B
thermocouples and the spray quench. The alterations, current configuration, and compromises
needed to move forward with testing will be discussed in this section.
F.1 B Thermocouples
Five pairs of thermocouples (a total of ten B thermocouples) are embedded in an axial line
along the north-west side of the reactor with wires exiting each of the five axial ports. An image
with the dimensions of the reactor is shown in Figure F-1. For referencing the ten thermocouples,
the following nomenclature was devised. The B thermocouples located closest to the burner are
position 1. The next closest to the burner are position 2 and this pattern is followed for 3, 4, and 5.
For each pair of B thermocouples, the thermocouple that is closest to the inner refractory wall is
“inner” and the other located further radially out is “outer”. Each pair of thermocouples were
spaced a radial distance of 50.8 mm apart with the inner thermocouple placed 114.3 mm from the
center of the reactor.
APPENDIX F. RECONFIGURED SUBSYSTEMS
126
Figure F-1. Dimensions of the POC’s combustion chamber [8].
The thermocouple pairs are designed for heat flux measurements. The objective was to
place the thermocouples within a section of refractory with the same thermal resistance as un
undisturbed section of the reactor, and therefore, the temperature difference of the thermocouple
pair would represent the same temperature difference found in any other azimuthal location at the
same axial distance from the burner. The thermal conductivity of the UltraGreen 45 (SR) refractory
is shown in Equation (F-1). Where Tmean is the average temperature of the outer and inner B
thermocouples at a given axial location. Using the temperature and thermal conductivity the radial
heat loss can be calculated for cylindrical coordinates.
127
𝑘𝑘 �
𝐵𝐵𝑅𝑅𝐵𝐵ℎ𝑟𝑟 ∗ 𝑖𝑖𝑖𝑖 ∗ 𝐹𝐹
�𝑈𝑈𝑤𝑤𝑜𝑜𝑟𝑟𝑚𝑚𝑈𝑈𝑟𝑟𝑚𝑚𝑚𝑚𝑚𝑚
= 23.98𝑅𝑅[𝐹𝐹]𝑚𝑚𝑚𝑚𝑚𝑚𝑚𝑚−0.0415
144 (F-1)
An image of a pair of two-holed ceramic tubes like those used in the reactor are shown in
Figure F-2. .The bottom tube contains thermocouple wires with a bead on the end representing
what the bead is like that is embedded in the refractory. The wires pass through the double bored
ceramic tubes exiting the outer radius of the refractory. On the outside of the refractory, insulated
thermocouple extension wires are connected to the B thermocouple wires as shown in Figure F-3.
to ensure the wires do not touch the metal port or flange.
Figure F-2. An example of the B thermocouple design used in the reactor. Two-holed ceramic
tubes spaced 50.4 mm apart. On the left end of the bottom ceramic tube an exposed bead is
visible.
128
Figure F-3. Insulated wire exiting the refractory inside the “two-inch” port of the reactor.
These insulated wires are connected to a ceramic terminal blocks as shown inside the 5.1
cm (2 in) port on the right side of Figure F-4. The electrical signal must be carried out of the
pressurized reactor to atmospheric conditions. To accomplish this, four pressure NPT feedthroughs
were used. The feedthroughs are mounted in four “1/8 in” NPT tapped holes as shown on the left
side of Figure F-4. A copper wire connects the feedthrough to the ceramic plate.
Figure F-4. Machined blind flange with the pressure NPT feedthroughs (on the left) and a
ceramic terminal block (inside the flange on the right).
129
On the outside of the reactor copper butt connectors connect the pressure feedthroughs to
the B type thermocouple extension wire. A pressure feedthrough and copper cut connector are
shown in Figure F-5. This extension wire carries the signal back to OPTO where the signal is
processed.
Figure F-5. “1/8 in” NPT pressure feedthroughs (on the right) carry the B thermocouples
electrical signal outside the pressure reactor. Copper butt connectors (on the left) used to connect
the pressure feedthroughs copper wire.
After the first time heating the reactor to its operating temperature and letting it cool to
room temperature; four of the B thermocouples stopped functioning. To determine if the
thermocouples were still functional, the blind flanges and ceramic tiles were removed. This
allowed for resistance measurements to be taken across the B thermocouples wires only. Four of
the ten thermocouples had infinite resistance and were open circuits. For this reason, it was
assumed that the four thermocouples were damaged somewhere within the refractory. The
damaged B thermocouples are 2-inner, 2-outer, 3-inner, and 4-inner. This means heat loss cannot
be calculated from thermocouple pairs 2, 3, and 4. For them to be replaced, the refractory would
need to be bored out and new refractory laid in its place with new B-Thermocouples installed. Due
to the time and cost that replacing the B-Thermocouples would take it has been decided to operate
130
the reactor for the near term with six of the ten B-Thermocouples operational and two of the five
heat loss calculations.
F.2 Spray Quench
The spray quench system designed by Skousen [9] introduces water into a refractory lined
reactor outlet upstream of a stainless-steel pipe section leading to the heat exchanger. The spray
quench water is needed to cool the exit gases to temperatures acceptable for stainless steel. Flue
gases exiting the combustion chamber have been measured as high as 1,316oC (2,400oF) while
operating near steady state with a firing rate of 70 kW. The stainless-steel connector’s maximum
operating temperature is 503oC (938oF) at 2.76 MPa (400 psi).
An atomizing spray nozzle is mounted in the center of the reactor exit pipe which is fed by
a high pressure water pump [9]. Skousen [9] tested the pump and nozzle system in air at room
temperature but not in hot exhaust gas with coal particles. When operated at high temperature and
with coal particles present, the upstream side of the nozzle became fouled and plugged. Even
though the nozzle material should have survived temperatures during heat-up where fuel without
particles was used, the nozzle would no longer function properly following overnight heating. The
high cost of the nozzle ($260) and the large lead time for replacement led to a search for a new
nozzle.
Additionally, the gas temperature downstream of the nozzle was not being reduced as
expected. One possible reason was the wide spray angle of the nozzle producing impingement on
the pipe wall and wall wetting instead of evaporative cooling. The nozzle is in the center of a pipe
with a 76.2 mm inner diameter. Water exiting the spray nozzle with an angle of 90o has 72.6 mm
before reaching the wall. Because of wall wetting a smaller spray angle is desirable.
131
The spray quench subsystem has a maximum pressure of 13.9 MPa (2,000 psig). The
nozzle must also be able to operate at more than 2854 kPa (400 psig). This will allow water to be
introduced at any reactor operating pressure.
The Bete Microwhirl (MW) 695 was selected to be the new atomizing spray nozzle based
on its ability to operate under the criteria identified above. The new nozzle has a spray angle of
70o, operates at lower pressures, and has a higher flow rate capacity than the previous nozzle.
OPTO has been programmed to control the temperature of the stainless-steel surface
thermocouples to be 316oC. With a fan blowing on the outside of the connector, the nozzle can
cool the surface to 316oC when the reactor is operating at 80 kW and has reached near steady state.
Although a higher temperature material was available (nickel alloy C-276) 316 stainless steel
nozzles and filters were selected because of the low cost ($37 compared to $200) and availability.
It was hypothesized that particles on the upstream side of the nozzle were created by boiling
water, creating mineral deposition from the culinary water, which were essentially, hard water
deposits facilitated by boiling. Although low flow rates of water were sufficient to keep the
stainless-steel connector from overheating, they were not sufficient to keep water in the nozzle
from boiling and rapidly creating a plugged condition in the spray orifice. For this reason, the
pump should always run when combustion products are exiting the reactor. When this rule has
been followed, the nozzle has not plugged.
132
G.1 First Attempt to Burn Coal
The first attempt was on March 18, 2021. That morning, the coal feeder was filled with
coal to the point it was overflowing out of the top. When the reactor reached 1338 KPa and the
coal hopper was at 1475 KPa the solenoid that separates the two systems was opened. The two
systems’ pressures did not equilibrate. There was evidence of coal being transported to the reactor
based on measured wall temperature increases, but combustion appeared to only last for a couple
of seconds. The coal hopper was vibrated using a hammer to assist in braking up any bridging, but
the reactor thermocouple temperatures never increased, and the camera never showed a flame. The
pressure differential between the hopper and the reactor suggested there was a plug in the tubing
between the two vessels.
It was discovered after depressurizing that filling the feeder to the top had clogged the filter
on the vent line isolating the vent line from the hopper. The only pressure transducer for the coal
feeder was on the vent line, downstream of the filter and due to the filter being clogged the pressure
in the coal hopper may not have been represented by the pressure transducer. The hopper and the
reactor may have been at the same pressure when the valve was opened creating a blockage. This
result introduced a new safety issue. The hopper pressure must be known and controlled and
therefore an additional pressure transducer was installed on the pressure relief line directly
APPENDIX G. PRESSURIZED COAL TESTS
133
connected to the hopper. It was also determined that the coal feeder should not be filled to the top
in future experiments.
After depressurizing, the distributer that controls the coal fluidization process was
examined. A tuyere that was press fit into the distributer had fallen out and coal had passed through
the hole to the upstream side of the distributer and the large hole would allow CO2 to tunnel through
the coal without fluidizing it.
The four tuyeres were tack welded into place on the distributor to prevent them falling out
again. Also, it was determined that moisture from the hopper hydrostatic test may have still been
present or may have produced corrosion residue that could block coal flow through small openings.
G.2 Second Attempt to Burn Coal
The second coal test occurred on April 30, 2021. The preheating and pressurization process
were followed as described in the March 18 test. The coal used to fill the hopper on March 18 was
left in the barrel without the lid on the barrel. This coal was used to fill the coal feeder for the
second coal attempt.
The solenoid that separates the coal feeder and reactor was opened twice and evidence of
coal being transported was apparent for a few seconds after each time the solenoid opened. The
coal hopper was hit with a rubber hammer to produce a vibration with the intent of breaking up
any bridging or plugging of coal in the hopper, but this action did not produce coal flow. It was
concluded that the coal left in the barrel had accumulated moisture and caused coal to become
sticky preventing fluidization and flow. To avoid this, on future attempts the barrel should be filled
the morning of the test from the coal bags and the coal bags should be closed at the end of each
test. After removing the coal from the feeder, the tuyeres were inspected and found to be in place
and minimal coal was found upstream of the distributor.
134
G.3 Third Attempt to Burn Coal
The third coal attempt occurred on May 14, 2021. This was the first coal attempt that
resulted in coal being transported continually for an extended period. The hopper was filled with
15 kg of coal. Initially, the coal hoper would not pressurize because the pneumatic ball valve at
the top of the hopper would not seal shut. After five minutes of leaking, the ball valve suddenly
sealed, and the coal feeder was able to pressurize. This issue was resolved in future tests by making
sure the ball valve was free of coal dust prior to each test.
The reactor was pressurized to 1475 kPa and the coal hopper was pressurized to 1549 kPa.
At this point, the solenoid separating the two systems was opened allowing coal to start flowing.
The Horiba oxygen dropped to zero indicting fuel rich conditions. The flow rate calculated from
the hopper load cell forces was fluctuating between 14 and 18 kg/hr for about 20 minutes. An
attempt was made to increase the oxygen flow rate to match the coal flow and complete
combustion, but the oxygen flow rate would not increase above a total of 23 kg/hr between the two
oxygen mass flow controllers. The POC camera showed a flashing intermittent bright luminous
flame. The radiometer on the top window (level 1) indicated a signal but the radiometers on the
other axial locations showed no signal. It is not known if the lack of signal was due to a flame that
did not extent beyond the first level or because the windows became covered with unburned coal
and char blocking the radiative intensity from reaching the radiometers.
After twenty minutes of flow, the next fifteen minutes showed the coal flow slowing until
it completely stopped. The hopper pressure began to increase above that of the reactor pressure.
After the pressure rose to 34 kPa (5 psi) above the reactor there was a sudden increase in
temperature within the reactor and a decrease in hopper mass indicating the coal had plugged and
that this pressure differential pushed open the plug. to The test continued for about thirty minutes.
135
Although the fluidization flow was higher, the coal flow rate was lower. The flow eventually
decreased, and it was assumed the hopper was empty. The mass indicated by the hopper load cells
ended at 2.75 kilograms.
G.4 Fourth Attempt to Burn Coal
The fourth pressurized coal test performed on May 21, 2021 is the first to accomplish
sustained periods of coal combustion burning in total approximately 20 kg of coal. The objective
of the test was to vary the flow rates of fluidizing CO2 and vent CO2 and observe the resulting
impacts on coal flow rate while measuring wall temperatures, radiometer signal, reactor pressures,
temperatures and flow rates.
A three-hour period encompassing the starting and ending period of coal flow is shown in
Figure G-1. Before analyzing the results, it is important to understand the time response of the
various measurements. The coal feed rate is measured from the change in weight of the coal bin
over time. Because of the change in mass of the coal is small compared to the mass of the coal
hopper, it was estimated that it would require thirty seconds of feeding coal to register a measurable
change in coal mass. The data shown in Figure G-1 are averaged over 2.5 minutes creating a delay
between a change in coal feed rate and a measurable signal indicating the change. Another
challenge with the coal feed rate is that the coal bin was periodically impacted with a rubber
hammer to reduce bridging. This would cause a brief but large force on the transducers resulting
in a false positive or negative indicated flow of coal. In order to remove these false signal, all data
suggesting greater than ±50 kg/hr of coal feed for a 30-second period were removed and replaced
with the average from the previous 30-second period.
136
Figure G-1. Coal mass flow, CO2 flow rates, radiometer signal and level 1 inner refractory
temperature as a function of time for pressurized coal test #4.
Conversely, the data for radiative heat loss, has a time response of less than a second and
therefore rapid fluctuations in the data can be considered real and the data were not averaged or
137
smoothed. The fluidization flow rate was measured based on the response of thermistors within
the mass flow controllers with a response time on the order of seconds. The vent CO2 flow was
recorded by hand within a minute of the change and therefore the timing of a change in vent and
net CO2 flow rates are only within 1-2 minutes of the real timing of the event. The temperature
obtained from B-Thermocouples has a response time on the order of minutes, but the
thermocouples respond faster than the refractory in which they are mounted and can therefore be
considered to provide a correct temperature with time although the refractory temperature has a
delayed response relative to the coal flow and radiometer signal due to its large thermal mass.
Looking at the data as a function of time, a correlation can be seen between the level 1
refractory temperature, the radiative intensity signal, and the coal mass flow rate. Each of these
are indicators that coal was flowing and burning in the reactor. For example, the peak in
temperature seen at 12:00 pm indicating the end of a heat addition period coincides with the end
of a period of fluctuating radiometer signals and the end of a period of high coal flow rate. That
peak in temperature is then followed by four smaller peaks in temperature which can also be seen
to coincide with brief periods of radiometer signal and coal flow.
It was expected that coal flow rate would correlate with net CO2 flow rate. This
correlation cannot be seen in the data. There are several instances when coal flow increased with
decreasing net CO2 flow, the opposite of expectations. There are periods where CO2 flow slowed
and stopped without a change in the net CO2 flow. Because the net CO2 flow is the carrier of the
coal to the reactor and must be related to coal flow, the data suggest other variables were also
impacting the net CO2 flow rate beyond the fluidization and vent CO2 flows.
The four peaks in temperature occurring after 12:00 pm and the associated coal flow
resulted from closing the valve between the coal hopper and reactor and opening the valve after
138
building a small pressure difference of about 14 kPa. This produced temporary flow but the flow
then stopped on its own without any changes to the CO2 flow rates.
After several attempts to create steady flow, and after banging on the hopper to remove
the possibility of bridging as a problem it was decided to end the test. A steady stream of coal
exiting the bottom of the hopper after the end of the run also suggested the flow was not stopped
due to bridging. It was hypothesized that CO2 could be tunneling or channeling through instead
of fluidizing the coal.
A brief period of 45 seconds of radiometer data from the same test is shown in Figure
G-2. In these data, the baseline voltage of the radiometer at the start of coal flow was subtracted
from the data. This allows the relative magnitude of the coal flame radiation to be seen. These
data indicate that the coal is burning in a series of high frequency radiative events that are likely
caused by uneven coal flow. Between most events, the radiative flux returns to a low value near
zero, perhaps that of the background wall emission. The data were recorded at 1-second intervals
which appears to be too slow to capture each individual event. This is consistent with
observation from a digital camera located to view at level 1 opposite the radiometer. The signal
for level 2 was flat, indicating that either the flame was not reaching level 2 or the window had
become opaque with unburned coal. Post-test inspection of the windows suggested that they
were opaque from unburned coal produced during periods of high coal flow that exceeded
stoichiometric. The amount of oxygen fed was limited by a broken regulator upstream of the
mass flow controllers which limited total mass flow.
139
Figure G-2. A selected period of high radiometer signal at level 1 early in the test.
G.5 Fifth Attempt to Burn Coal
The fifth attempt occurred on May 28, 2021. The primary purpose of this test was to test
the hypothesis that the coal was not feeding because of tunneling and poor fluidization. As a result,
a new fluidization mechanism was used on this attempt. A 316 Stainless-steel, wire mesh (325 x
2300, McMaster-Carr P/N 9419T39) was used to evenly distribute the fluidization CO2 through
the coal. Two, 3.2 mm (1/8 in) thick ANSI class, 300 Garlock, gaskets with bolt holes, McMaster-
Carr P/N 9472K676, were used to seal the mesh between “two-inch” flanges. The mesh was cut
to 7.62 cm (3 in) diameter circles so when placed concentrically with the gaskets the outer edge
would be halfway between the gaskets inner circle and the inside of the bolt pattern. RTV sealant
was applied to the perimeter of the mesh to adhere it to the gaskets. This prevented the mesh from
shifting with respect to the gasket during installation. Although it was found during testing that
140
coal did not pass through a single layer of mesh, two layers of mesh were installed between the
Garlock gaskets.
During this attempt to burn coal only eight minutes of coal burning was achieved. The
morning of the test, the coal hopper was filled with 35 kg of coal. The reactor was pressurized, and
the mass flow controllers were set to the desired setpoints (except oxygen). Like the fourth attempt
to burn coal, the oxygen regulator, when fully opened, only provided a maximum pressure on the
downstream side of the regulator of 2.0 MPa (280 psig) although there was an upstream pressure
of over 2.9 MPa (400 psig). As in the fourth attempt to burn coal, this limited the oxygen mass
flow controllers to produce a combined maximum flow of 23 kg/hr. As shown in Table 5-4, the
desired combined oxygen flow rate needs to be 30.1 kg/hr to achieve 3% O2 in the flue gas.
With the coal feeder at 28 kPa (4 psi) higher than the reactor, the solenoid separating the
two systems was opened. This smaller pressure differential caused a smaller initial pulse of coal
to be transported and initially, the windows were not covered in coal. At 8:26:00 the CO2
fluidization flow rate was 1.7 kg/hr. After half a minute of operation, the window at level 1 (back
side) was entirely covered in coal preventing the POC camera from viewing inside the combustion
chamber. It was also later discovered after shut-down that coal was deposited on the front and back
windows at level 2. After two minutes the camera at window (level 1) was still covered in coal and
nothing could be seen in it. This excess of coal in the system is evidence that the reactor was
running overall rich.
Because the oxygen introduced to the burner could not be increased the coal mass flow rate
was decreased by decreasing the fluidization CO2 flow rate. At 8:28:45, the fluidization flow rate
was decreased to 1.5 kg/hr and then at 8:29:21, the fluidization flow rate was again decreased to
1.07 kg/hr. From 8:30:50 the fluidization flow rate gradually decreased until it reached its
141
minimum flow rate of 0.8 kg/hr at 8:33:00. The mass flow controller could not decrease the flow
rate lower than 0.8 kg/hr. This resulted in the coal mass flow rate decreasing and the excess coal
on the window was burned away. However, flow rate stopped all together. At 8:35:17 the
fluidization flow was gradually increased until it reached the originally flow rate of 1.7 kg/hr, but
the coal flow rate did not return to the original flow or even register a flow. The solenoid was
closed to allow a pressure differential to build up and the solenoid was opened to try to get coal to
flow again. The coal flowed for a few seconds after opening the solenoid, but the coal feeder
started increasing in pressure independently of the reactor and a plug was believed to have formed.
The pressure differential continually increased and at a pressure difference of 20 psi the solenoid
was closed and the systems depressurized. After depressurizing, the mesh was inspected and was
found intact. Minimal coal was found on the upstream side of the mesh.
A plot comparing the coal flow rate during attempt #5 using the mesh and the steadiest
flow observed in attempt #4 using the tuyeres is shown in Figure G-3. The coal mass flow rate for
both attempts was backward averaged over 30 seconds and the flow rate is plotted every minute.
The entire attempt #5 coal feeding lasted about 12 minutes were as the attempt #4 had three
sections of coal feeding in the reactor for a total of # minutes. The most constant 12 minute section
of attempt #4 feeding is shown in the plot below. Although it is difficult to draw a conclusion from
a 12 minute period of feeding coal, the tuyere fluidization mechanism, used in attempt #4,
oscillates 6 kg/hr every three minutes compared to little or no oscillations at this frequency for the
mesh design used in Test #5. Additional data needs to be collected before it can be concluded that
the mesh produces a more uniform flow than the tuyeres but this initial data is promising.
142
Figure G-3. Calculated coal mass flow rate during the fifth coal test with a mesh fluidization
mechanism.
Figure G-4 shows radiometer signals (levels 1, 2 and 5) and the coal mass flow rate. The
radiometer signal is backward averaged over 1 minute with error bars representing the standard
deviation of 60 signals and therefore represents the magnitude of signal fluctuations. Each
radiometer’s initial raw voltage signal is between 1.0 to 1.3 volts when looking at a room
temperature surface. To compare the changes in radiometer signal relative to their initial signal
each radiometer’s raw voltage signal before feeding coal, the signal at 7:00:01 a.m., was subtracted
from each radiometers respective sampled raw voltage signal. This allows the change in signal
produced during the period when coal was being fed to be shown.
143
Figure G-4. Radiometer voltages during the eight minute coal burn of the fifth coal test with the
mesh fluidization mechanism.
Initially when coal was first fed, high fluctuations in the radiometer signals are recorded.
High fluctuations from the bottom radiometer shows that the flame is capable of reaching the level
5 port when running under rich conditions. The magnitude of the fluctuations at level 5 begin to
decrease after 8:32 when the coal mass flow rate signal also begins to decrease. This is evidence
that the pulses of coal being fed are reaching level 5 less frequently and that there is a sufficient
oxygen to enable complete combustion.
A similar trend is seen from the level 2 radiometer. The magnitude of fluctuations decreases
shortly after the coal mass flow rate starts to decrease but the fluctuations do not stop completely.
This is evidence of the flame was still reaching level 2 but more intermittently as a greater number
coal pulses reached complete combustion prior to level 2. The level 1 radiometer continually sees
144
the flame and this is represented by the consistent magnitude of radiometer signal fluctuations
even when the coal mass flow rate is decreasing.
The radiometers raw voltage was sampled and recorded every second. One minute of the
radiometer signal is shown in Figure G-5. The period shown in this figure was selected because at
this time the coal was beginning to feed into the reactor and the viewing ports were not yet covered
in coal. High frequency fluctuations in radiometer signal are evidence of unsteady coal flow to the
reactor. It is important to note that during this attempt it was observed that the coal would feed to
the reactor in pulses. The level 1 signal pulsed continually with a frequency of 3 to 4 seconds. This
is similar to the frequency that was notice in the attempt 4 data (Figure G-9). The level 1 signal
pulsing is evidence that the coal consistently reached the viewing port. The level 2 radiometer
oscillated less frequently meaning the flame only occasionally reached the level 2 viewing port.
The level 5 radiometer signal did not pulse at all, suggesting complete combustion was achieved
before the level 5 viewing port. The negative value suggests that the radiometer was viewing, at
least in part, the hot walls opposite the viewing port and that these wall were slightly cooler at this
time than they were when the offset was subtracted prior to the introduction of coal.
145
Figure G-5. Radiometer signal oscillations for one minute during attempt 5.
In conclusion, the mesh fluidization mechanism appears to perform better at eliminating
the medium 1-3 minute fluctuations in coal feed but does not change the high frequency pulsing
that is on the order of 3 seconds. However, only eight minutes of coal data were captured during
this attempt due to what is believed to be a coal plug. The plug could not be cleared by increasing
the pressure of the coal feed 138 kPa above the reactor pressure. Another problem was the oxygen
capping out at 23 kg/hr. This preventing the system from running overall lean for a coal mass flow
rate at or above 100 kW.
G.6 Sixth Attempt to Burn Coal
The sixth attempt to burn coal occurred on June 18, 2021. The purpose of this attempt was
to gather more data on the performance of the mesh fluidization mechanism. Prior to the test, the
146
oxygen regulator was repaired, enabling a higher pressure for the mass flow controllers to produce
sufficient oxygen to run the reactor overall lean even though the coal firing rate exceeded 100 kW.
A plot of coal mass flow rate and radiometer signal is shown in Figure G-6. The solid lines
represent a 2-minute running averages for the coal and radiometer signals. The voltage recorded
on the radiometer just prior to the start of coal feed was subtracted from the radiometer so that the
relative change in signal from the start of coal combustion could be evaluated. The error bars
represent one standard deviation of the data. For coal flow the standard deviation was calculated
over 2 minutes the radiometers was also calculated over 2 minutes. The magnitude of the error
bars represents the relative fluctuations in coal flow rate and radiometer signal.
Figure G-6. Coal mass flow rates and radiometer signals for the duration of pressurized coal Test
#6. The data shown is averaged over 2 minutes.
147
At 9:00 am the solenoid was opened between the coal hopper and the reactor allowing coal
to flow. The coal flow rate increased rapidly and remained flowing until 9:50 after which it stopped
temporarily. It was revived by hitting the hopper with a rubber hammer at 9:50. Following this
event, it was assumed that the coal had bridged in the hopper and to avoid bridging again, the
hopper was hit with a hammer every five minutes. At 10:15, the coal flow stopped again and was
temporarily revived, but the hopper mass was 3 kg (almost empty) and it was assumed that a steady
coal flow was no longer possible. During the 40 minute period from 9:10 to 9:50, three dips in the
coal mass flow rate are seen after which the flow rate returns to approximately 15 kg/hr. The
average coal flow rate over this period was 15.0 kg/hr with a standard deviation of ±2.9 kg/hr.
Radiometer absolute values increase at all three levels (1, 2 and 5) immediately following
the solenoid opening and correspond with the rapid rise in coal flow rate. Fluctuations in flow
rate are also seen in the first 10-15 minutes between 9:00 and 9:15 am. This is a period when the
coal flow exceeds 15 kg/hr and the oxygen flow rate was 36 kg/hr the overall fuel air mixture
was likely fuel rich. While the absolute voltage levels de not decrease until the coal stops
flowing at any of the levels, the magnitude of the fluctuations at levels 2 and 5 decrease
significantly after the first 15 minutes. This indicates that a fluctuating coal flame is not seen
consistently below level 1 after the initial surge in coal flow rate. The fluctuations in radiometer
signal match visual observations of camera images of the flame at level 1 and indicate the
presence of a pulsing flame. The flame can be seen at level 1 throughout the test at all coal flow
rates but only at level 5 when the fuel air mixture is overall rich. Under these fuel rich
conditions, unburned coal is likely entering the stainless-steel connector and accumulating on the
heat exchanger tubes and in the condensate drain port.
148
The refractory inner wall temperatures are shown in Figure G-7. The temperatures are
increased during the two regions of continual coal flow mentioned earlier. These regions of
temperature increase being between 9:10 and 9:50 and 10:00 to 10:20, a few minutes after coal
feed is registered above, indicating a slight lag between when coal begins to burn and the refractory
begins to heat up. Other than these two regions the refractory temperatures decreased. After natural
gas heat-up the reactor’s bottom thermocouples are hotter than the top thermocouples. After five
minutes of introducing coal, the temperature of the top of the refractory exceeded that of the bottom
refractory. This is evidence that the coal introduced to the reactor quickly burns producing a short
flame and a small region where most of the heat release occurs.
Figure G-7. Refractory inner wall temperatures calculated from B-type thermocouples during
coal feeding.
149
The refractory heat loss is shown in Figure G-8. In this figure the heat loss is approaching
its maximum at approximately 9:40. The coal and natural gas heat losses are assumed to have
similar shapes. Therefore, the coal heat loss must decrease after reaching its maximum before it
reaches steady state. Since the coal has not approached steady state, conclusions cannot be drawn
about the coal heat loss.
Figure G-8. Refractory heat wall calculations calculated from B-type thermocouple during coal
feeding.
A higher resolution of the radiometer data for level 1 is shown in Figure G-9. The
fluctuation in voltage appears to cycle with a period of between 2-3 seconds. These fluctuations
are consistent with those observed during coal tests 4 and 5 where the tuyeres were being used.
The level 2 radiometer signal is relatively constant and is evidence that the flame is not reaching
the second port during this time of the test.
150
Figure G-9. The oscillation of the Level 1 radiometer during coal attempt 6.
Critical exhaust system temperature measurements are shown in Figure G-10. The
stainless-steel connector was cooled with the spray quench internally and a fan externally. The
spray quench was allowed to introduce water continually from the beginning of the heat-up
process until the end of feeding coal. A new Bete spray nozzle P/N PJ 20 was used for this test.
This spray nozzle used has a higher water flow capacity and a larger spray angle (90o) than the
Bete spray nozzle used in previous coal attempts. The atomizing mechanism for this nozzle is
also different. This nozzle uses impingement to atomize the water whereas the previous nozzle
used a critical orifice.
151
Figure G-10. Stainless Steel connector surface K-type thermocouples during coal feeding.
The performance of this Bete nozzle did not have any noticeable differences from that of
the MW695 Bete nozzle. The maximum recorded surface thermocouple temperature was more
than 150oC lower than the maximum allowed operating temperature of 503oC. Although the
connector was still heating up by the end of the test, there was still 150oC of thermal capacity
available before reaching the connected would reach the maximum allowable temperature.
Flue gas temperatures at several locations are shown in Figure G-11. The flue gas
temperatures were relatively constant during the coal burning period. The heat exchanger exit flue
gas temperature was constant at the supply water temperature. By performing an energy balance
and assuming the mole fractions of the products of combustion are 0.83 CO2, 0.12 H2O, and 0.05
O2. The 1450 K temperature at the exit of the reactor suggests that from the 122 kW pressurized
coal flame 42% of the energy is lost in the reactor. The thermocouple’s recorded temperature, 1450
152
K, is not adjusted for the radiative loses and therefore the loses in the reactor are likely less than
42%.
Figure G-11. Flue gas K-type thermocouples during coal feeding.
By assuming the heat loss of the reactor is the average of the maximum B thermocouple
heat losses at level 1 and level 5 a heat removal of approximately 17 kW or 14% of the 122 kW
can be estimated in the first 1.8 m (6 ft) of the reactor. This is significantly lower than the estimate
based on the reactor exit thermocouple. Some of the difference in these two values could be
attributed to the heat loses of the flue gas as it travels from the level 5 B thermocouple to the exit
reactor thermocouple and the inaccuracy of the reactor exit thermocouple due to radiative losses.
A model presented in Carpenter’s thesis [8], Figure 4-4, shows the CFD predicted temperature
profile along the length of the POC reactor for each of the three burner design cases. During
attempt #6 the reactor was operated closes to Case 3. This model predicts that the reactor exit
153
temperature would be 2000 K. The model therefore predicts a heat loss between 42% and 14%.
Additional work would be required to measure this heat loss more accurately.
In conclusion, the mesh fluidization mechanism was successful in feeding coal to the
reactor for 40 minutes of continual feeding. Although the reactor quit feeding, it was easily revived
to the same mass flow rate by banging on the hopper and fluidizing region. This suggests that after
40 minutes the coal had bridged. With occasional banging, the system was able to feed the entire
22 kg of loaded coal. A vibrator should be investigated as a means of producing continuous flow.
The rebuilt oxygen regulator allowed sufficient flow to run the reactor overall lean when
introducing coal at a firing rate of 15 kg/hr. The oxygen mass flow controllers also had the capacity
to introduce more oxygen if desired. Additional coal attempts are needed to fully characterize the
reactor for longer continual coal feeding.
top related